Skip to end of metadata
Go to start of metadata

RHQ 4.8.0

This is the RHQ 4.8 release. It was released on June 25th, 2013

Installation note
RHQ 4.8.0 no longer has a GUI-based installer and its installation routines have changed even since 4.7. Even if you are familiar with installing earlier versions of RHQ, be sure to see Installing the Server for updated installation and upgrade instructions. Specifically, you are now required to use the rhqctl Control Script to install (and start) the RHQ server components.
Upgrade note
If upgrading from RHQ 4.2 (4.3 and later are not affected) you must first make a manual change to your database. Have a look at the first note on the RHQ 4.4 release notes
Please note
RHQ 4.8 is recommended with Java7, and also supports Java6.

New Features and Changes (since RHQ 4.7)

Cassandra backend for metric storage

RHQ 4.8 is the first release that uses Cassandra as backend store for metric values. This has been done to increase scalability. Numeric metric data is no longer stored in the RHQ relational database. There is no option or support to still store numeric metric data in the RHQ relational database. The use of Cassandra as a data store is intended to be an implementation detail to the greatest extent possible. This has some significant implications.

  • RHQ will manage the Cassandra storage nodes (This release introduces some initial management support. More management support will be added in subsequent releases).
  • An RHQ agent will have to be run along side Cassandra.
  • RHQ will not support using an arbitrary Cassandra installation for its storage node.

Some of the critical cluster management functionality is still undergoing development and testing; consequently, this release only supports a single node cluster. Please review the following bugs for known issues.

  • BZ 976814 - Cassandra username/password cannot be changed during before installation
  • BZ 976822 - Cassandra and RHQ Storage plugins log UnsatisfiedLinkError to stderr
  • BZ 951993 - The RHQ Storage Node installer has a dependency on com.sun classes
  • BZ 960726 - Cannot switch to a JVM that does not support snappy compression after the Storage Node is installed
When running rhqctl start on OS/X or IBM jre it is possible that a message like the following show up - this is harmless.

REST API

GUI

  • Graphing enhancements: New date range navigation control and ability to drag a range over a graph. Video (4 min) Date Range Slider Controls
  • Availability Charts redesigned
  • Portal Charts now have large and small versions (to fit narrow column portlets)
  • Monitoring --> table tab changed to 'metrics' in preparation for new consolidated metrics tab (in 4.9.0)

Server

Plugins

Cassandra Plugin

A new Cassandra plugin compatible with Cassandra 1.2.x is part of this release. The plugin supports monitoring and configuration for features exposed via Cassandra's JMX interface. Additional support for some important internal configuration settings (Eg. gc_grace_seconds) is also available. The plugin has enhanced support for taking and restoring snapshots.

Known Issues

Deprecations

  • Method org.rhq.core.util.exec.ProcessExecutor#redirectStreams has been deprecated. Callers should switch to using a new method org.rhq.core.util.exec.ProcessExecutor#redirectAllStreams method.
  • REST-API Endpoints introduced in 4.5.1 to create platforms and resources have been deprecated - replacements are provided

API changes

  • The return type of method org.rhq.enterprise.server.measurement.MeasurementDataManagerRemote#getAggregate was changed from org.rhq.core.domain.measurement.MeasurementAggregate to org.rhq.enterprise.server.measurement.MeasurementAggregate. It is the same class, only package is different. CLI scripts should remain compatible with this change, for Java clients depending on EJB remote API, fixing the imports and recompilation will solve the issue.

Translations

The GWT part of the UI has partially been translated into German, Portuguese, Japanese, Chinese, Czech, Russian and Korean. The language should be automatically selected depending on your browser settings. You can explicitly access other translations by appending a locale specifier in the URL. For example to select the German translation you would append ?locale=de to the base URL, e.g. http://localhost:7080/coregui/?locale=de.

Supported locales are:

  • zh for Chinese
  • cs for Czech
  • de for German
  • ja for Japanese
  • pt for Portuguese
  • ru for Russian
  • ko for Korean

Please ping us if you want to help translating the UI to your language. Translations are done via the translations project on GitHub, which also has some instructions on how to start.

Bug reporting

Please report all bugs you find in Bugzilla. If you find a bug that has been recorded in the above list, please leave a comment on them especially if this needs special steps to reproduce.

List of resolved Bugzilla entries

Please consult Bugzilla with a target release of RHQ 4.8.0 for a list of resolved issues

Download

You can download the release here

Commit log (since RHQ_4_7_0 )

  • Heiko W. Rupp <hwr@redhat.com> (77):
    • removed plugins
    • Revert "removed plugins"
    • Add another email for Thomas
    • Code cleanup and improvements for the REST-api
    • BZ 960529 - check for non-existing resources when obtaining availability
    • Make this test use the RHQ server, so that no external EAP needs to be set up.
    • BZ 960959 apply inventory status filtering with COMMITTED as default for searches.
    • BZ 961787 - try to determine AS/WFLY via domain api version before falling back to directory names, as this is more reliable.
    • Provide the old method signature as fallback in case other plugins that inherit from this one, still need it.
    • BZ 961656 - enable LDAP support for the REST-api as well.
    • Fix a small typo
    • We don't need standalone-osgi.xml
    • Disable for the moment
    • BZ 962858 check passed resource id and return 404 if invalid
    • Print the failing class along with the failed field.
    • BZ 962853 - make getRresourceId return an int instead of a string.
    • Enable paging on (some) endpoints that return lists
    • Support paging in the body with the appropriate media type
    • Support paging in the body with the appropriate media type
    • BZ use explicit paging fields, as there is no default. In the specific case, the 'name' field was not unique and thus 'random within the resources with the same name'
    • Bit of cleanup and testing of alert->ack and alert->purge
    • Add some translations
    • Print the failing class along with the failed field.
    • BZ 962853 - make getRresourceId return an int instead of a string.
    • Enable paging on (some) endpoints that return lists
    • Support paging in the body with the appropriate media type
    • Support paging in the body with the appropriate media type
    • BZ use explicit paging fields, as there is no default. In the specific case, the 'name' field was not unique and thus 'random within the resources with the same name'
    • Bit of cleanup and testing of alert->ack and alert->purge
    • Add some translations
    • Allow to use the .jsonw prefix to request the application/vnd.rhq.wrapped+json type
    • BZ 966381 - allow to import servers and update resources.
    • BZ 963691 - return scheduleId as int.
    • BZ 967448 - keep the prefix on redirect and also return alert definitions as XML
    • BZ 970517 - return a message that LDAP users need to register first.
    • BZ 970593 do not prepend the links with /rest
    • BZ 970592 make sure resourceId and scheduleId are rendered in decimal
    • Make the docs generator understand the notes field on @ApiOperation
    • Use the avg-sign as "Durchsch." is too long
    • BZ 967832, 967744, 967742, 967488, 967450 Various fixes and improvements in handling of AlertDefinitions
    • Allow for the notes to be XML - e.g. docbook snippets.
    • BZ 963804, 963667 improvements to the metric endpoints
    • BZ 835647 allow to specify a configurable number of buckets for metrics instead of the old default of 60.
    • BZ 929203 Allow to upload plugins via REST api.
    • BZ 962855 If the 'validate' parameter is passed, invalid ids are flagged.
    • BZ 972572 - Set the correct Cassandra default port.
    • A few more DE translations
    • Small docs update.
    • BZ 972657 - add a missing dependency to jboss logging
    • Update pom versions
    • Mention the data migrator after install and provide it in the server downloads section.
    • Put the data migrator notice to the upgrade command
    • Use the correct script to stop pre-48 servers.
    • Allow to pass a different C* data dir. Fix agent upgrade.
    • Allow the upgrade script to run the data migrator. This is currently postgres only.
    • There is no need to reserve that much memory upfront
    • Use port 8786 for debugging, as 8787 is used by the server.
    • Make the storage upgrade wait a little to prevent port conflicts with the old storage that may take some time until it finishes stopping.
    • In order to wait for storage to stop, check if the process is till available.
    • Make the graphs the default subtab for metrics again.
    • BZ 965201 show the condition value as a column in recent alerts lists. Remove the extraInfo= part, as it is confusing.
    • Minor cleanup
    • Fix a typo
    • Check for empty group definition expressions and negative recalc intervals
    • BZ 949659 Cope with an alert plugin having an empty <alert-configuration>
    • Get this to compile again and bump the JRuby version.
    • Add a description for the operations endpoint and mark it as experimental.
    • BZ 970970 hide traits in the table that are disabled to be orthogonal with the metrics.
    • BZ 965342 - make sure the alert definition is valid. From Elias Ross
    • Add the creation of an overview report.
    • Allow to search for resource types by name
    • BZ 918051 - Don't allow to pass arbitrary names for the resource.
    • Add a link to the resource from the alert details view.
    • BZ 949065 Prevent a NPE if Events show up before the EventManager is fully initialized. Reviewed by mazz.
    • BZ 965350 Prevent NPE if the server plugin manager is not yet started.
    • Some translations for the new chart time selectors.
    • Name the new graphs subtab "Graphs"
  • Hudson <jboss-qa-internal@redhat.com> (1):
    • tag RHQ_4_8_0
  • Jay Shaughnessy <jshaughn@redhat.com> (90):
    • BZ 921927 Work on search bar issues: - Add RecordDoubleClick handling to SearchStrategies, allows us to impl SavedSearch delete via DblClick. - Allow TableFilter to be disabled by SearchBar impl, allows SavedSearch mode to not trigger searches. - Add SavedSearch updateName/updatePattern/delete - Fix up keypress delay by canceling timer as needed - Add SavedSearch hover showing pattern in suggest list - Tie suggestion box size/positioning to search text box - Add some more ignored keys - A variety of other behavioral tweeks - Remove unncessary field defs in search strategy impls
    • BZ 921927 - add confirm box for delete - add some tooltips - clean up some related message properties
    • Bug 846451, 949048, 949062 * Users with just Resource Group access cannot access alerts on their Groups * Failure adding member to group with alert definition * Removing member from group with group alert definition removes all instances of the alert definition
    • BZ 949082 RFE: Allow alert ack/delete from recent alerts portlet and reporting view
    • BZ 921927 Prevent concurrent searches for suggestion list entries or saved searches. Timing issues could cause duplicates to show up in the list.
    • Update some alert def tests to better test for bug 949048 and 949062
    • BZ 872731 Platform utilization data fails to be retrieved if agent is slow or down and results in many UI errors Server returned FAILURE with no error message
    • BZ 950461 Update serialVersionUID for criteria classes
    • Bug 857438
    • Bug 872731
    • BZ 921927 Tried to make text boxes resize when browser is resized but failed to get it to work. I tried to emulate http://www.smartclient.com/smartgwt/showcase/#layout_form_spanning but it didn't seem to work within the Toolstrip. Instead, to better support narrow browsers I: - reduced the size of the search text box by 100 px - removed the unnecessary leader spacer of 40px - was able to fully hide the savedSearch text box space when not being used
    • BZ 911652 Redundant metrics displayed when selecting resource group metric on dashboard portlet
    • BZ 906385 [RFE] trait change alert conditions should support value expression
    • Fix test
    • Fix typo in property name
    • Add -Pdev copy of jars to dev-container
    • - handle executor command lines for windows - use windows service for hosting cassandra - use tanuki java service wrapper and introduce rhq-storage.bat, analogous to rhq-server.bat. And rhq-storage-wrapper.conf. - change rhqctl.sh to rhqctl - log exception cause for RHQControl - add rhqctl.bat and rhq-storage-installer.bat for windows , to match rhqctl and rhq-storage-installer.sh - storage installer changes for windows and some fixes - some fixes to rhq-server.sh|bat scripts - come .classpath changes for eclipse
    • More eclipse tweeks
    • Fix property name typo
    • The cassandra log filename was huge, I don't think we want the whole path in the name, do we?
    • this is a manual cherry-pick of a master commit addinga convenience method to the ASClient.
    • fix a small issue in the rhq-installer.bat
    • Working towards rhqctl install support on windows...
    • On windows default the cassandra data to HOMEDRIVE:/rhq/storage/...
    • Fix issues in install marker file placement.
    • Get Agent to install and start as windows service via 'rhqctl install'
    • More rhqctl work - update start/stop/status commands for windows compat - fix logdir mkdir issues in rhq-server.bat - add option to ControlCommand.getCommandLine() to not append .sh to the script name - add Remove command (although currently not sure if we want this in rhqctl, so commented out)
    • Fix 'rhqctl status' for windows, avoid throwing exception on aganet check
    • A few eclispe/build/master-merge tweeks
    • Dev-container work - remove rhq-autoinstall.sh|bat and its use
    • Remove ccm-cli as it is no longer needed for dev-container support
    • Extend -Ddbsetup/dbreset to the storage db (cassandra). This provides more robust support for the dev env in that now the RDB and C* are handled together.
    • Add initial StorageNode entity support: schema/dbupgrade/entity/criteria
    • Server needs more permgen, even in dev container
    • more .classpath changes
    • Change the cassandra heartbeat quartz job to be non-volatile because volatile is treated as non-volatile in a Quartz cluster, and was generating logging at 5s intervals. Also, just move the job creation into the existing scheduleJobs method.
    • Add Server Plugin for injecting factory-installed alert templates for RHQ. This will be used to create pre-defined storage alerting, alhough at the moment it just creates a dummy template for alerting when the rhq agent goes down.
    • just some name changes to distinguish templates and maybe make things more clear.
    • Fix issues in DiscoveryBossBeanTest that could cause test interaction problems. - Incorporate test class name into plugin/entity naming - use unusual ids - delete entities based on unique naming, not in ways that could wipe another test class's data
    • Add the actual Process to the ProcessExecutionResults and ProcessExecutorResults. This allows the caller to inspect or detroy the process if necessary.
    • - fix spelling of artifactId: rhq-cassandra-ccm-arquillian - add ProcessExecution.addArguments() which is needed for Windows, as setArguments() will blow away the internally added [] args.
    • Make sure we always use forward slashes when generating the dev-container rhq-storage.properties file. Backslashes are treated as escapes by Properties.load().
    • eclipse: update stax cassandra driver version
    • Add support for server itests on Windows. Arquillian+Surefire+Cassandra+Windows did not like the way we used the suite deployment extension as a hook for starting and stopping a test cassandra cluster. The spawned cassandra process seems to confuse things such that the arquillian process and its deployment (eap) process would hang, and the surefire report would not be generated. Likely due to the fact that Windows does not deal with "background" processes the same way as linux. This was after already trying to trim the test cluster from 2 nodes to 1 node, for similar process-related issues.
    • turn off default use of FilePreferences for agent wrapper
    • Fix discovery issues
    • add rhq storage plugin to ear
    • Fix artifact name for storage plugin
    • - Fix issue with storage node setup - Go back to using pid files to kill cassandra processes on linux
    • trivial - fix type in log message
    • remove placing the dep plugins in the jar lib. The useClasses feature will get the classes from the previously loaded plugins as needed.
    • CassandraNodes (and StorageNodes) can't defer to JmxDiscoveryComponent.upgrade() because the resource keys are not JvmResourceKeys. Override upgrade() and just return null, there is no resource upgrade, these are new resources.
    • Fix test failure issues caused by unclosed streams. Also, make an anal retentive change to StreamUtil to make absolutely sure we clean up the streams when closeStreams is true.
    • Fix test that started failing after the change to remove JOIN FETCH from Criteria calls that include paging and optional data. It seems the new approach changes slightly the way the data is returned.
    • Fix test that started failing after the change to remove JOIN FETCH from Criteria calls that include paging and optional data. It seems the new approach changes slightly the way the data is returned.
    • Change helper method from public to private. For classes annotated with @Test all public methods are treated as tests so you have to be careful.
    • Use only apache logger and not the jboss logger from the base class.
    • BZ 966665 - Criteria queries with paging and sorting on non unique fields may return same item on different pages
    • Use only apache logger and not the jboss logger from the base class.
    • BZ 966665 - Criteria queries with paging and sorting on non unique fields may return same item on different pages
    • remove use of method not supported by GWT
    • remove use of method not supported by GWT
    • Make sure sub-queries used for updates do not include implicit paging/sorting.
    • BZ 966665 - Criteria queries with paging and sorting on non unique fields may return same item on different pages
    • remove use of method not supported by GWT
    • Add new JMX metric for "VM Memory System:Heap Usage Percentage". This provides a way to monitor/alert actual memory consumption as it is calculated as HeapUsage/HeapCommitted.
    • Use new JMX plugin metric "VM Memory System:Heap usage Percentage" to provide a factory installed alert template for an RHQ storage node. The actual numbers probably need tweeking, currently its HUP > 75% with 10 of 15 dampening.
    • Get rid of the unnecessary and mispelled local method AlertDefinitionManagerBean.createAlertDefinitionAndRerurnIt(). It's value was that it returned the created alert def as opposed to just the id of the alert def. So, instead, have AlertDefinitionManagerBean.createAlertDefinitionInNewTransaction() return the AlertDefinition as opposed to the id.
    • Fix an issue with alert template application to discovered resources. The root cause has to do with sending attached entities across Tx boundaries:
    • Make sure our injected cassandra vm alert template has its relevant metric enabled and reported frequently.
    • Fix test cleanup and add a bit more to test code.
    • eclipse .classpath changes
    • add more info to exception
    • Add a little more help to rhqctl
    • Fix problem setting session timeout for non-settings-manager users. This was related to a past fix for BZ https://bugzilla.redhat.com/show_bug.cgi?id=857438 allowing GUI session timeout to be configurable.
    • BZ 915291 ] Errors in server log after first attempt to login with LDAP user
    • Fix issue with alert duration job when the resource in question is removed from inventory during the duration period.
    • rhqctl work - add some feedback to the install process while waiting for server install - improve some messaging - switch to using dmr client (as opposed to log scanning) for install completion checking - improve -Pdev update of dev-container the server-control module to update the entire module
    • Rhqctl work: add the 'remove' command to remove services when on windows. Add initial 'upgrade' command. Does nothing other than validate args at the moment.
    • On windows, make 'rhqctl install' perform a windows service reinstall even when the component is already installed. This is more like the legacy behavior for install/remove on windows.
    • prevent upgrade if RHQ is already installed
    • rhqctl install/upgrade work - move default server/storage agent dir out from under server so it can survive updates and not invalidate associated RhqAgent resources. - refactor Install and Upgrade to share abstract base class - variety of other use case work
    • Change to dev-container structure to push the server under dev-container/rhq-server. This allows us to install the agent in the new fixed location, which is as a sibling to the server directory, in a location that does not have to change as server versions are upgraded.
    • BZ 974876 [RFE] ResourceUpgradeFacet should support upgrade of plugin config properties - Implement PC config upgrade for StorageNode resources so that paths can be update to handle RHQ upgrades and maintain the same resource. - Add test code to ResourceUpgradeTest
    • Update StorageNode plugin to use new resource upgrade feature to update plugin config prop directory paths after an RHQ upgrade.
    • add a new rhqctl debug config for eclipse
    • Fix some upgrade bugs when upgrading from a non-rhqctl version (windows only issues)
    • More Upgrade fixes - Allow --storage-config option for upgrade. This is valid only for the case when upping from pre 4.8 to support the SN install. - Fix Windows issue when copying over old key/trust store settings - Enhance the help text a little for the pre 48 options
    • Fix upgrade issues - fix agent upgrade of existing agent in the default location (i.e. no --from-agent-dir specified) - On Windows, ensure exit codes are passed back from the storage installer and rhqctl scripts. The former is very important as otherwise errors in the storage installer do not prevent an attempt to upgrade/install the server.
    • Beef up RUN_AS/RUN_AS_ME option for the server and storage windows services. Make sure the RUN_AS options can be used without hanging rhqctl on a password prompt. These are now like the agent, although by default RHQ_STORAGE_PASSWORD_PROMPT is set false for storage and server, since they are primarily installed via rhqctl. It still defaults to true for the agent-wrapper, since agents are primarily not installed by rhqctl.
  • Jirka Kremser <jiri.kremser@gmail.com> (1):
    • Merge pull request #1 from lzoubek/master
  • Jirka Kremser <jkremser@redhat.com> (92):
    • Adding "/cassandra/" to .gitignore. ("./cassandra/" didn't work for me on Fedora.)
    • NPE was thrown by the installer when the node was not running. Adding the check for null when closing the cassandraSchemaManager.
    • BZ 876132 ] - Remote agent install form requires the ins. path to be non-empty, but the tooltip says it can be empty - Tooltip has been changed. Also the whole view for this use case was improved.
    • BZ 876132 ] - Remote agent install form requires the ins. path to be non-empty, but the tooltip says it can be empty - Allowing blank installation path. It should search for the agent install directory if the path prefix is provided or to choose a default one.
    • BZ 923458 - Versioning for dependencies and plugins not consistent - version of liquibase set to 2.0.3 (there was a CNF exception when running itests-2)
    • Bug 846451, 949048, 949062 * createAlertDefinition() renamed to createAlertDefinitionInNewTransaction() (itest-2 module didn't compile)
    • BZ 823908 ] - User should be redirected to Operations->History tab after operation execution on autogroup - The getListViewPath method slightly modified. Now it takes into consideration the case when scheduling the operation for autogroups.
    • BZ 824439 ] - Links in 'Group: Operations' and 'Group: alerts' portlet are not working in 'Summary->Activity page of autogroup - URL of links has been fixed for operations portlet.
    • BZ 824439 ] - Links in 'Group: Operations' and 'Group: alerts' portlet are not working in 'Summary->Activity page of autogroup - URLs have been fixed for alert portlet. Based on EntityContext.isAutogroup().
    • BZ 857438 ] - Create a way to let GUI session timeout be configurable - Adding the GUI Timeout as a system config property.
    • BZ 950461 ] - Update serialVersionUID for criteria classes - serialVersionUID++
    • API checks: adding a record to intentional api changes. (restriction of visibility of serialVersionUID static field defined on class SavedSearchCriteria)
    • Restriction of visibility of serialVersionUID static field defined on class SavedSearchCriteria.
    • BZ 824439 ] - Links in 'Group: Operations' and 'Group: alerts' portlet are not working in 'Summary->Activity page of autogroup - URLs have been fixed for events portlet, resource metric portlet and resource's Alerts->Definition table column (when the alert was originaly defined on the autogroup). The ResourceTreeView component now works all the time (there was an issue when it wasn't possible to select any node in the tree and user had to refresh the page if the URL of autogroup was directly passed to browser or accessed via any link in the UI).
    • Adding 3 methods for resource/type (un)ingoring to intentional API changes, namely: ResourceManagerRemote.uninventoryResourcesOfResourceType(), DiscoveryBossRemote.unignoreAndImportResources(), ResourceTypeManagerRemote.setResourceTypeIgnoreFlagAndUninventoryResources()
    • Adding 2 classes to the intentional API changes, namely: DisabledResourceComponent and DisabledResourceDiscoveryComponent. These two classes were introduced on Jan 25 for as a hot fix. Now they are not used anymore, because the ignoring of a whole resource type solves the original issue.
    • BZ 951266 ] - d3 Charts on Inventory -> Summary -> group metric portlet don't work properly for various groups - There was an timing issue with using the count down latch in class CompositeGroupMultiLineGraphListView. Also, different criteria ("filterVisible" in particular) have to be used for fetching the data for autogroups.
    • Splitting the JavaScript code generated by GWT compiler into smaller files. It reduces the time needed to load the login page.
    • Changing the "activeByDefault"for Oracle mvn profile to false.
    • Preventing the timing issue with rhq.cassandra.seeds property by explicitly calling the jboss-cli and setting it for the AS.
    • moving the .propSet away from the /home/jkremser/workspace/rhq/dev-container//bin directory; fixing a typo in StartupBean
    • Adding the executable bit for rhqctl script.
    • Moving the RMIContextFactory class to a separate maven module, because of the dependency on com.sun.* package. When depending on this Maven module, adding the exclude tag not to propagate the dependencies (on com.sun) transitively.
    • Adding the dependency to cassandra-driver-core and rewriting the ClusterInitService.waitForSchemaAgreement() using the CQL.
    • Replacing the Thrift code with JMX calls, this includes: refactoring of StorageNode; cahnges to ClusterInitService; Thrift port was replaced with JMX port
    • Turning off the Thrift-based rpc by default.
    • Adding criteria finder for StorageNode entity and itest for the finder.
    • Adding local interface for StorageNodeManager, moving the criteria finder from TopologyManager to StorageNodeManager, following the same file structure for itests
    • Adding the dependency to core-domain for module rhq-installer-util.
    • Removind SuppressWarnings and making the call type safe.
    • New icon for storage node entity.
    • Adding a comment explaining why simple String.format() can't be used.
    • Fixing a typo in a sample CLI script.
    • Making the height of the content of the sections in the navigation on the left hand side automatic, depending on the number of items it has.
    • Adding fetchResource() method to StorageNodeCriteria class.
    • Adding new data source and data source field class for Storage Node entity.
    • New views for Storage Node entity (list view and detail view); also adding the new criteria finder to the TopologyGWTServiceImpl.
    • Fixing a typo.
    • Adding a few sample operations for Storage Node, namely (Start, Stop, Restart, Start/Stop RPC Server, Decommission).
    • The name of operation for shutting down the storage node is called "shutdown", not "stop".
    • BZ 958169 ] - Modified alert definition is not saved on confirming the save message after clicking on 'Back to List' button - Making the "Back to List" button disabled when editing the alert definition.
    • BZ 958169 ] - Modified alert definition is not saved on confirming the save message after clicking on 'Back to List' button - I18n (Instead of "Edit", now the button label says "Edit Condition").
    • BZ 958169 ] - Modified alert definition is not saved on confirming the save message after clicking on 'Back to List' button - New method for creating an alert definitions was added to SLSB, because the UI requires the ids of newly created alert conditions.
    • Making the StorageNodeManagerBeanTest#testInit test passing; building the rhq-storage plugin by default
    • Just adding the cassandra plugins to plugin validator (validating the plugin descriptor, resource type, etc.)
    • Improvements to RHQ irc bot. New set of commands is now supported.
    • rhq irc bot: disabling the !pto and !support commands on Freenode channel
    • Exposing new metric for Cassandra plugin - Tokens. Also improving the jmx plugin to be able to interpret the measurements when the List is returned from attribute.getValue(). The meassurement value is then calculated as the number of elements in the obtained list.
    • Exposing new metric for Cassandra plugin - Ownership. This has to be calculated manually in the getValues() method on StorageServiceComponent.
    • Exposing new metric for Cassandra plugin - Ownership. Forgot to add one method.
    • Refactoring: moving the MeasurementAggregate from server/jar to core/domain in order to be able to reuse it from other modules
    • Fixing the build (missing import)
    • Adding the getLoad() method to the StorageNodeManagerBean. It is half-way don, few aggregate metrics still needs to be added. Also exposing the StorageNodeManagerBean in CLI interface.
    • just a typo
    • Making StorageNodeManagerBean.getLoad() method more efficient using named queries. Support for correct unit formatting of StorageNodeLoadComposite.toString().
    • BZ 967542 ] - UI shows confusing units when editing Availability Duration alert condition type
    • Improvements to RHQ irc bot. New set of commands is now supported.
    • rhq irc bot: disabling the !pto and !support commands on Freenode channel
    • BZ 967542 ] - UI shows confusing units when editing Availability Duration alert condition type
    • Merge branch 'master' into feature/cassandra-backend
    • Adding calculated metric heap percentage usage to the StorageNodeLoadComposite.
    • Fixing bug in IRC bot (isRedHatChannel); pto command now shows PTOs longer than 1 day; ensuring the bot do not notify users when using the pto and support commands
    • Adding some localization for Storage Node entity UI.
    • fixing typo in my previous commit (3d3f1aae4)
    • Ending curly bracket was missing.
    • Bash completion script for rhqctl command.
    • Fixing the rhq-master-api-check jenkins job. Only harmless changes were introduced by commit 85f75e0
    • RHQDomain GWT module now takes into consideration properties gwt.locale and gwt.userAgent. It decreases the compilation time in for the dev profile (provided only 1 user agent and 1 locale is set).
    • Adding JavaDoc to StorageNodeManager.
    • GUI improvements for StorageNodeDetailView
    • Annotating the change in MeasurementDataManagerRemote as intentional. For more details see the justification in intentional-api-changes-since-4.7.0.xml.
    • Making storage node JMX and CQL port read-only in UI.
    • Enabling heap used metric for RHQ storage node by default.
    • GUI improvements for StorageNodeDetailView
    • Updating the rhqctl bash completion script. Also improvind the path completion for parameters that require paths.
    • Support for the upgrade command in rhqctl autocompletion script.
    • BZ 973140 ] - Editing dialog for alert conditions does not load condition parameters
    • BZ 973746 ] - metric definition id column is named schedule_id - Replacing the "Schedule ID" with the "Definition ID".
    • Fail fast, if the StorageNode has no associated resource.
    • Preparing UI for expandable rows. Prototyping with some artificial data.
    • Adding a new metric for Cassandra plugin. This metric is a calculated one and it denotes the percentage usage of the partition where the data files of C* are stored, if there are more of them, the maximum is used.
    • Making the getLoad method more safe. The entityManager.merge(node) wasn't doing the right thing.
    • Improving the StorageNode UI, making the rows expandable. Also adding the operation history view as a section to the SN detail page.
    • Logging the root cause of the problem in CriteriaQueryRunner.initialize() when debug LVL is on.
    • Renaming "DiskSpaceUsedPercentage" metric to "Calculated.DiskSpaceUsedPercentage" on C* plugin.
    • Adding the Calculated.DiskSpaceUsedPercentage aggregate metric to the StorageNodeLoadComposite and modifying the StorageNodeManagerBean.getLoad method to load this information.
    • Adding the new aggregated metric from StorageNodeLoadComposite to the StorageNodeLoadComponent (the ListGrid with the populated data).
    • Pretty formatting for rhqctl status command.
    • Setting up the operation mode for storage node to avoid the DB constraint violation (null is not allowed) when running StorageNodeManagerBean.scanForStorageNodes() after clean build.
    • Making StorageNodeManagerBean.scanForStorageNodes() safer (in order not to fail with DB constraint violation).
    • BZ 974436 ] - Edit alert condition -> metric value is not converted to target units - Converting correctly the value based on it's measurement units family.
    • Updating the rhqct bash completion script.
  • John Mazzitelli <mazz@redhat.com> (69):
    • BZ 872731 - this fix adds the ability to define a timeout for a specific remote POJO obtained by the client factory.
    • Squashed commit of the following:
    • BZ 535289 - by default, don't get schedules from criteria if resource is not committed this addresses a problem when the group metric table view aggregated metrics that included ignored resources
    • fix problem with the goal - verify doesn't appear to be a valid goal name, check is valid.
    • BZ 535289 - when ignoring a resource, it is appropriate to set their availabilities to UNKNOWN since we will not be tracking their avails anymore. this commit adds an API to AvailManagerBean to allow us to explicitly set resource avails. When ignoring resources, we set them to UNKNOWN.
    • BZ 849711 - make the warning before deleting agents more scary
    • BZ 960265 bring back the GWT calltime page
    • BZ 817825 don't show "Loading..." for leaf nodes in resource type drop down menus this still doesn't fix the Platform plugin's platform type submenus.
    • BZ 535800 - allow the agent to download the agent update binary over the configured SSL channel
    • BZ 535800 - can't use default protocol because it is automatically initialized and cached, not allowing us to use a customized one add a main() for quick testing
    • BZ 951382 - if keystore/truststore is left undefined in agent config, we'll explicitly set the pref for the user using the "conf/" location
    • BZ 903402 - fix code if windows drive is missing but otherwise the path is absolute
    • BZ 699893 911692 - if the user ran EAP 5 with cmd line option -Djboss.server.data.dir=/some/data/dir then the AS5 plugin should look there, too, to try to find the jnp-service.url file
    • provide some additions to the sample skeleton plugin add a new "skinny platform" plugin sample, used mainly for testing - provides a minimal plugin you can use to run an agent with
    • provide another sample plugin that demonstrates using the embedded plugin extension model
    • BZ 951382 - for upgrades, let's use the agent preferences upgrade feature to set the cert store locations explicitly and copy the files from data/ to conf/ which is the new default location. Note that if customers already explicitly set the locations of their cert stores, we leave those intact and continue using what the customer asked to use. We only do anything IF those settings weren't set and thus defaults were taking effect.
    • BZ 957282 957288 add bundle target of "Base Directory"
    • BZ 963982 - make sure we persist the token when the registration was successful, even if the server endpoint was bad
    • add jboss logging to eclipse classpath
    • BZ 962841 962845 fix the resource type admin pages to support types with multiple parents (i.e. types with <runs-inside>)
    • revert this back - closing the streams is fine here, useless to try to close the inner buffered stream when closing the outer stream will also do that too
    • BZ 967990 the installer now turns off the admin console
    • add jboss logging to eclipse classpath
    • BZ 962841 962845 fix the resource type admin pages to support types with multiple parents (i.e. types with <runs-inside>)
    • BZ 968357 535831 get IBM JRE to run on both server and agent with SSL/certs
    • BZ 966160 bind some internal services to local loopback only
    • BZ 970230 - remove ExampleDS from RHQ Server during install
    • Merge remote-tracking branch 'origin/master'
    • * do not maintain a second agent-config .xml file for fear we will go out of sync over time (as proof, we've already diverged from the main agent xml copy). * do not clean the top rhq-agent pref node - that will purge everything and could effect other agents running with non-default nodes (we use this for agentcopy/agetnspawn for example) - just purge the rhq-agent/default node since we'll install using that as the default * fallback to the agent to determine its default server binding address to use - this allows the agent to do its default behavior (which uses the "rhqserver" DNS alias or, if that is unknown, using the local canonical host as a fallback) * do not purge the security token, if one exists - this lets us (poorly) do upgrades as well as do a clean install - problem is if its an upgrade, we should still keep all settings. We should consider not doing clearAgentPreferences() in org.rhq.server.control.command.Install
    • fix more agent install stuff for rhqctl * you can now override any agent pref via --agent-preference, including agent token if you want. you can specify more than one. * removed unused option * slurps in the config file (whether custom or the default one) and populates the preference node with it, but still maintaining security token if one exists. * sets the setup flag to true, always, since the agent is never able to ask setup questions anyway via rhqctl
    • BZ 951382 forgot to bump this number up in the config xml
    • * rhqctl no longer supports putting agent and storage installs in custom locations via agent-dir and storage-dir- they always go in a well-known location under the root isntall dir * --storage-data-root-dir now allows you to set the root directory where all the storage data dirs go - this allows one to easily install even if you aren't running as root * the installation will fail if the custom --storage-data-root-dir is not empty * fixed the wording on some of the help text to make it more clear about things (i.e. installing storage will imply installing an agent too; if you install just a --server, lets the user know they should have a storage node somewhere else)
    • turn of auto-update of the agents that are installed via rhqctl; these agents are managed by rhqctl
    • fix help text - we don't support agent in console right now
    • if one of the data directories is not empty, we abort the storage installer since this indicates we are probably stepping on another storage node installation
    • trivial tweeks to help text
    • remove some obsolete code that isn't used
    • provide an option to install, but not start, the agent. This provides a window of opportunity for users to reconfigure the agent before starting or add things like keystore/truststore files to it
    • abort the server install if something is already listening to port 9999 (which is probably another EAP instance that is in the way)
    • switch from using protected fields to protected getter methods.
    • first stab at upgrading rhq-server.properties. needs testing
    • * copy the old oracle jdbc driver (if one exists) to the new server install * make sure we do not overwrite the db during the upgrade by setting the autoinstall.database value to auto
    • log a user-friendly error to tell the user what he should do if the storage node can't write to the data directories
    • chmod +x the rhqctl script
    • fix the error message to show the option name prefixed with – to make it clear what it is referreing to fix the chmod +x for rhqctl
    • fix copy-n-paste bug
    • do some checking to make sure the build actually has the EAR where it can be accessed
    • throw exception immediately if the EAR URL could not be obtained
    • fix more things in the rhqctl upgrade - always ensure the autoinstall.enable is true and get the oracle module from the old server properly
    • more work to get agent and server upgraded. still nothing done yet wrt storage upgrading. this stuff still needs work. need to know if we should upgrade server and storage (i.e. if they weren't installed, we shouldn't upgrade)
    • * if anything is installed already (including just the agent) the upgrade will not continue. An upgrade will only happen if you unzip a distro and not install any component (i.e. you must upgrade, not install) * the old server install rhqctl script is invoked with "stop" argument so any old components still running are stopped prior to the upgrade continuing. * the agent is started after the agent upgrade happens, unless the user told us not to start the agent
    • stop the agent using the agent wrapper script if a from-agent-dir was specified
    • * change help text for --storage-data-root-dir to explain when you can and can't use it * rename method getPre48RhqServerScriptName to getRhqServerScriptName since that script name still exists and is still the same name even in RHQ 4.8+ * abort the upgrade if --storage-data-root-dir is specified but you are not upgrading a 4.8+ install. This is only valid when upgrading an older version that did not have a storage component.
    • fix the for-loop so it doesn't rely on the order of files coming back from listFiles
    • during server upgrade, make sure we copy over any keystore/truststore files that were in the old server. We put them in the new server's standalone/configuration directory.
    • rename --agent-no-start argument to --agent-start and have its boolean value be optional.
    • change the name of the --agent-start option to --agent-auto-start ... we'll come to a consensus soon on the name of this arg
    • fix the help text - /var/db is not the location
    • reorder some things so we dump the info message saying "stopping components" prior to actually stopping them. also, don't stop the agent until we are sure we can proceed with the upgrade.
    • make sure merging server properties from older RHQ 4.5.1 and earlier is done properly.
    • make sure we look at the different default file locations specified in older servers' properties file
    • make sure we get the oracle driver even in the case of updating an older JBossAS 4.2.3 based RHQ release
    • remove unused constant. make the "private static" keyword order consistent. OCD meter went off the charts when I saw "private static" then "static private"
    • fix stopServer so it stops the server
    • do not mention "pre-4.8" or any version number in help/error messages since it won't be helpful to JON users
    • the default location for the agent is the parent dir of the --from-server-dir
    • fix comments in rhqctl
    • there is no example.storage.properties, remove its reference from help text
    • BZ 976882 - do not copy the old plugins to the new agent; the new agent needs to pull down the new plugins from the server before it can start the PC for the first time.
  • John Sanda <jsanda@redhat.com> (459):
    • Initial commit for cassandra-bundle module
    • Install rhq cassandra bundle on server start up.
    • Refactoring logic for embedded cluster deployment
    • Periodically check that we can connect to at least one cassadra node
    • Need to parameterize the listen_address and rpc_address properties in bundle recipe
    • Changing deployment property cluster.type property to is-embedded
    • fix dev profile to copy artifact to correct location
    • Start nodes after they are installed on the file system
    • Install schema after starting first node
    • Include JNA libraries in cassandra bundle and make logging configurable
    • adding work around for snappy-java library with Java 7
    • Initial commit for cassandra-common-itests
    • rename module directory to cassandra-common
    • Initial commit for server-metrics module
    • fixing cassandra version
    • Initial commit for MetricsServerTest
    • Initial commit for ClusterInitService
    • using ClusterInitService to wait for cluster initialization
    • Porting code from github repo for calculating aggregates
    • fixing calculateAggregatesForOneScheduleWhenDBIsEmpty test
    • Merge branch 'rhq-on-as7' into feature/cassandra-backend
    • Porting more code for data point queries and tests.
    • Renmaing EmbeddedDeployer to BootstrapDeployer
    • moving classes to package org.rhq.cassandra
    • Initial commit for CassandraClusterManager
    • use CassandraClusterManager to shut down the cluster.
    • Adding a ping method to ClusterInitService
    • adding numNodes attribute to @DeployCluster annotation
    • Merge branch 'rhq-on-as7' into feature/cassandra-backend
    • update container build to create cassandra jdbc driver module
    • First pass at generating jdbc driver and data source configuration
    • Merge branch 'rhq-on-as7' into feature/cassandra-backend
    • Include cassandra related properties in rhq-server.properties in dev builds
    • Adding initial support to specify cassandra hosts during installation
    • update start up logic to first look at system properties for cluster configuration
    • upgrading cassandra bundle to cassandra 1.2.0-beta1
    • removing beans.xml since CDI does not work yet
    • refactoring ant packaging scripts
    • adding num-tokens and ring.delay properties to bundle recipe
    • do not use dashes in bundle property names
    • Updating test to set new deployment properties
    • initial commit for dbsetup.cql
    • first pass at porting code to use cql for reading/writing metrics
    • Moving logic out of DeploymentOptions constructor and into new load method
    • (cql port) update metrics index when inserting raw metrics
    • renaming table from metrics_work_queue to metrics_index
    • (cql port) Adding back logic to purge schema between test runs
    • (cql port) first cut at aggregating raw data
    • initial commit for MetricDAO and MetricDAOTest
    • simplifying the findMetricsIndexEntries method
    • adding method to update metrics index
    • adding method in MetricsDAO to insert and retrieve raw metrics
    • refactoring MetricsServer.addNumericData to use MetricsDAO
    • delegate to MetricsDAO to update metrics index
    • adding methods to MetricsDAO to insert and retrieve aggregates
    • (cql port) adding 6 hour table back to schema
    • Adding method to find aggregates by date range
    • (cql port) adding back support for producing 6 hr aggregates
    • refactoring code to update metrics index
    • adding option to skip cluster shutdown during test runs
    • CassandraClusterManager should not automatically purge cluster directory
    • adding 24 hour metrics table back into schema
    • adding method to purge index for metrics table
    • purge one hour metrics index after calculating raw aggregates
    • adding support for setting column meta data when inserting raw metrics
    • upgrade bundle to cassandra 1.2.0 beta2
    • fixing whitespace error in query and debugging tests
    • move cluster init logic out of test and into CassandraClusterManager
    • cleaning up tear down logic
    • Removing IdGenerator as it is not needed
    • more updates to aggregation related tests to use CQL/MetricsDAO
    • minor refactor to remove unused method param
    • refactor findRawDataForContext method to use MetricsDAO
    • Merge branch 'rhq-on-as7' into feature/cassandra-backend
    • Replacing usage of jdbc driver with datastax native java driver
    • Adding more cassandra config properties to bundle recipe
    • adding config properties from previous commit to DeploymentOptions
    • fixing property names
    • Adding initial support for using MetricsServer in actual running server
    • adding back support for fetching aggregate data points via CQL
    • adding back test for fetching 6 hour graph data points
    • Removing Hector test util methods that are no longer needed
    • changing property name to be more consistent with other property names
    • set the TTL for 1hr, 6hr, and 24hr data
    • Merge branch 'master' into feature/cassandra-backend
    • adding initial support for authentication and authorization
    • parameterize cassandra username/password
    • adding cassandra username/properties to rhq-server.properties
    • add authentication to session initialization
    • add method to fetch raw metrics for multiple schedule ids
    • remove dead code
    • adding method to MetricsServer to retrieve raw data for a resource group.
    • adding intial support for fetching raw data for group from cassandra
    • add metrics profile for just building metrics related code
    • Merge branch 'master' into feature/cassandra-backend
    • Adding initial Cassandra impl for MeasurementDataManagerBean.getAggregate
    • Adding initial support for fetching aggregate raw data for compatible groups
    • Refactor findDataForResource to fetch data from cassandra
    • removing dead code
    • Adding methods to retrive latest metric value for a resource
    • Removing last usage of MeasurementDataManagerUtility in MeasurementDataManagerBean
    • add member instance of MetricsDAO
    • adding support for fetching aaggregrated group data
    • adding support for fetching summarized group aggregate
    • adding test for fetching group summary aggregate
    • refactoring duplicate logic into shared helper methods
    • forgot to include pom.xml when previously committing files for new module
    • initial liquibase integration for schema change management
    • adding dev profile to cassandra-schema build
    • adding script to build libraries from source that are not yet in maven repos
    • Merge branch 'master' into feature/cassandra-backend
    • use compact storage since we won't be using collections with these tables
    • upgrading cassandra bundle to 1.2.0-beta3
    • install source artifact for debugging
    • adding exclusion to filter out extraneous, overlapping dependencies
    • adding enabled flag to more easily enable/disable tests for debugging
    • fixing dev profiles to copy artifacts to new location under modules directory
    • fixing bad merge and adding cassandra dev properties
    • Initial commit for SchemaManager
    • removing dependency on the cassandra-schema module
    • fixing indentation
    • first cut at integrating cassandra into new installer
    • Updating installer logic to be consistent with what we do for the RDBMS
    • updating CassandraClusterManager to use the new SchemaManager
    • upgrading to cassandra 1.2.0-rc1
    • checking if schema exists before creating it
    • Removing dependency on hector
    • initial commit for jboss module to deploy embedded cluster for development
    • update script to build against C* 1.2.0-rc1
    • fixing dependency scope
    • forgot to commit maven assembly
    • adding support for updating rhq-server.properties when rhq-ccm runs
    • Renaming testng listener as part of refactoring for CassandraClustermanager
    • initial commit for the new CassandraClusterManager
    • initial commit for CLI which is the class that the rhq-ccm script now calls
    • major refactoring of the former cassandra-common module
    • Moving impl of cluster shutdown out of testng listener and into CCM
    • adding option to rhq-ccm.sh to shutdown the embedded cluster
    • fixing error from merge conflict
    • refactoring command execution code
    • adding start command to rhq-ccm
    • adding support for shutting down specified nodes
    • add support for starting specified nodes
    • parameterizing rpc_port property in cassandra.yaml for bundle deployment
    • fixing rpc_port related property names
    • exposing commitlog.dir property so that default value can be overridden
    • exposing saved.caches.dir property so the default value can be overridden
    • exposing log.dir property so that the default value can be overridden
    • exposing listen.address property so that default value can be overridden
    • more property clean up
    • removing intial.token property since we are using vnodes
    • major refactoring of BootstrapDeployer
    • Do not unpack bundle each node deployment
    • adding some docs
    • store the hostname, thift, and native ports for each seed node
    • adding comments in rhq-server.properties for changed rhq.cassandra.seeds property
    • removing raw table info from system settings view
    • adding windows support for rhq-ccm
    • add back statement to write installed file marker for embedded cluster
    • return nodes from createCluster even if the cluster already exists
    • Merge branch 'master' into feature/cassandra-backend
    • fixing module name in ant script output
    • use read-only urls for git repos
    • fixing logic for passing nodes to SchemaManager
    • use a specified tolerance when comparing doubles
    • commenting out metrics profile for now while debugging CI builds
    • only need to deploy cluster once for test suite
    • more CI build debugging
    • Revert "more CI build debugging"
    • disable jsp compiler to see if it resolves other CI build issues
    • bumping up version of jetty jspc maven plugin to debug CI build errors
    • more CI build debugging
    • jenkins complained about missing version
    • fixing jsp compile errors
    • using batch inserts for aggregate data and for metrics index updates
    • don't create empty batches
    • tell shrinkwrap to only use the local repo to resolve deps
    • get host names from rhq.cassandra.seeds system property
    • initial commit for rhq-ccm maven plugin
    • add support for running itests with cassandra
    • allow the cluster dir to be set via a system property passed to maven
    • uncomment tests
    • adding profile for building rhq-ccm maven plugin
    • Downloading: http://repo1.maven.org/maven2/org/codehaus/plexus/plexus-compiler-api/1.5.3/plexus-compiler-api-1.5.3.jadding support for deploying embedded C* cluster once from arquillian
    • adding log4j.properties for debugging CI issues
    • forgot to add arquillian module to build
    • have shrinkwrap only grab deps from the local repo
    • add back call to notifyAlertConditionCacheManager
    • move clean up logic into finally blocks to ensure db clean up happens
    • check to see if the schema exists in the maven plugin
    • using min and max for to calculate buckets for data points query
    • fixing test to read/write metrics from MetricsDAO and not from RDBMS
    • Merge branch 'master' into feature/cassandra-backend
    • updating OOB test to read/write metric data from cassandra
    • Merge branch 'master' into feature/cassandra-backend
    • adding an initial impl for doing OOB calculations
    • disabling test for now since it is no longer applicable with cassandra backend
    • comment out OOB calculations from control.jsp for now to get build working
    • add back call to calculateOOBs
    • make sure test uses same jvm as maven and the rest of the build
    • make sure cluster.dir property is set correctly
    • add tear down logic to clean up db
    • clean up bundle recipe so it no longer creates incorrect directories
    • print out java version to debug jenkins test failure
    • Revert "print out java version to debug jenkins test failure"
    • adding log4j.properties to get debug logging during test runs
    • trying to work around NPE occurring in test
    • Revert "trying to work around NPE occurring in test"
    • Revert "make sure test uses same jvm as maven and the rest of the build"
    • disable test until sigar issue is resolved
    • completely disabling JMXPluginTest for now
    • initial commit for cassandra-installer module
    • add a reminder for updating rhq-server.properties
    • add back jmx test to debug sigar issues
    • Revert "add back jmx test to debug sigar issues"
    • need to keep snapshot artifact names consistent with what is in module.xml
    • make jmx port configurable in cassandra installer
    • Merge branch 'master' into feature/cassandra-backend
    • upgrading to C* 1.2.1 and removing usage of liquibase for schema management
    • initial commit for metrics-simulator module
    • adding initial support for taking input json file to configure the simulator
    • making retention times and time slice durations configurable
    • add support for configuring data retention and time slice durations
    • adding support for running aggregation
    • update bundle recipe to make cassandra heap sizes configurable
    • adding the rest of the pieces needed to support making heap size configurable
    • making serveral changes to simulator to put a hurtin' on cassandra
    • adding some initial support for generating stats about raw data inserted
    • updating the statistics that are tracked
    • updating installer with cassandra 1.2.1.
    • fixing bug to reset counter for raw inserts per minute
    • only install/update the schema when it has not already been created
    • making cluster configuration and length of simulation configurable
    • disabling tests
    • need to configure compression on the client side now
    • more fixes for datastax driver upgrade
    • updating build script to build snappy-java
    • several updates to support using forked version of snappy-java
    • need to initialize the metrics configuration in MetricsServer
    • initialize DateTimeService with metrics configuration
    • inital commit for packaging simulator as stand-alone, cli app
    • more clean up for running the simulator from the command line
    • previously forgot to commit SimulatorThreadFactory
    • add defaults for simulation config values
    • add error handling to terminate simulation
    • some more clean up on the simulator
    • fix default cluster directory
    • cleaning up statistics sampling and reporting
    • do not use hard coded hostnames when creating session for simulator
    • Merge branch 'master' into feature/cassandra-backend
    • adding support for rhq storage node installer
    • updating parent version
    • initial comit for maven assembly and module.xml
    • add back C* deployment arquillian extension for server itests
    • fixing broken tests
    • fixing version mismatch on transitive dependency
    • verify node is up as part of the installation process
    • work around for https://issues.jboss.org/browse/AS7-2138
    • adding start and check-status options.
    • configure C* logging
    • trying to fix CI build
    • add options for setting commitlog, data, and saved caches dirs
    • add option for setting native transport port
    • make storage ports configurable
    • adding system dependency for rt.jar in attempt to fix CI build error
    • provide default schedule set when none are specified
    • initial commit for rhq uber script, rhq.sh
    • add support for installing the agent
    • initial impl for stop command
    • fix maven filtering and some clean up for rhq.sh start/stop commands
    • handle removing java prefs for agent on mac os x and redirect stdout for C*
    • update plugin to discover C* 1.2.x nodes
    • log warning messages if user does not have write perms to data directories
    • put cassandra plugin back in build and tell jmx plugin to ignore cassandra
    • include cassandra plugin in ear and make installer dir checking more robust
    • changing default settings and renaming installer to StorageInstaller
    • update error handling for directory checks
    • Merge branch 'master' into feature/cassandra-backend
    • remove dependency on bundles for cassandra deployer code
    • Removing EJB that is no longer needed since are not doing bundles-based deployment
    • fixing ear build
    • do not hard code cql port when initializing session
    • make stack size passed to C* JVM configurable
    • set the stack size property of DeploymentOptions to effect stack size change
    • do not block indefinitely when checking cluster initialization
    • upgrading Cassandra, Cassandra driver, and snappy-java
    • removing stuff to support SimpleAuthenticator and SimpleAuthorizer
    • make compression configurable in simulator
    • register shutdown hook at start of run method
    • adding logging
    • initial commit for server-control module which will replace rhq.sh uber script
    • initial support for stop command
    • initial support for start command
    • adding logic to start command to check if a service installed before starting it
    • forgot to include InstallerServiceImpl in previous commit
    • initial commit for status command
    • initial commit for console command
    • updating version of cassandra driver
    • need to build server-control before appserver
    • load snappy native lib using system loader
    • forgot to add osgi dependency which is needed to load snappy with native loader
    • reverting back to patched version of snappy-java
    • bump up version of cassandra driver
    • fixing logic to start/stop only that which is installed by default
    • update script to build the correct versions
    • Do not start the storage node if it is already running
    • update status command to check only that which is installed by default
    • update install command to check to see if service is already installed
    • add server-control module to default build
    • package and deploy datastax cassandra driver as jboss module
    • increase timeouts to allow enough time for cluster to initialize during tests
    • check that the service is isntalled before executing console command
    • updating rhqctl install command to check if service is already installed before proceeding
    • BZ 951180 ] Log error messages that Cassandra writes to stderr
    • provide a sufficient delay for the cluster to initialize before apply schema
    • deleting rhq.sh since it has been replaced by rhqctl.sh
    • wait for dev cluster to initialize during dev-container deployment
    • set the export flag to true to make cassandra driver visible to libraries in lib dir
    • refactor process execution code to not use java 7 APIs
    • make sure we store cassandra.pid in <RHQ_STORAGE_BASEDIR>/bin/cassandra.pid
    • increase delay for cluster initialization before running arquillian server tests
    • finally fixing the call to wait for cluster initialization
    • increase delay for cluster initialization
    • initial support for specifying alternate install directories for storage and agent
    • updating rhqctl commands to support alternate location for storage install
    • updating rhqctl commands to support alternate install directory of agent
    • adding initial support for calling batch scripts on windows
    • adding initial support for --server-config option for install command
    • make embedded cluster heap settings configurable for dev-container deployments
    • adding initial support for --agent-config option to install command
    • add validation for config options to install command
    • replace strings with constants
    • more merge conflict clean up that was reverted during rebase
    • adding initial support for storage-config option on install command
    • changing version of jspc plugin to match what we have in master
    • adding support for writing to and reading from rhqctl.properties
    • BZ 953712 ] do not return non-zero value on status command
    • BZ 952248 ] wait for cluster initialization
    • BZ 952781 ] Add runtime deps to plugin and add cql port to plugin config
    • display usage when command is invalid
    • forgot to include in previous commit
    • BZ 907485 956878] Configure storage node defaults to work with OpenJDK and IBM JDK
    • log the root cause when a command throws an exception
    • adding javadocs
    • minor change for command line path resolution
    • guard against NPE
    • BZ 907485 ] need to increase thread stack size on amd64 arch with openjdk 1.6
    • Merge branch 'master' into feature/cassandra-backend
    • updating work around for AS7-2138 for use with new jboss-eap-6.1 container
    • fix start command for agent and make rhqctl script executable
    • removing duplicate calls to chmod task in ant scripts
    • make sure rhq-storage.log is written to RHQ_SERVER_HOME/logs
    • add more detailed logging to install command
    • temporary fix for applying shcema changes when schema already exists
    • fix start command for server
    • adding new memory options to storage installer
    • return non-zero exit code when installer receives bad command line
    • fixing credentials for tests
    • fixing credentials for arquillian tests and reverting formatting for better readability
    • make sure default super user is used to create schema and fix deps
    • only execute C* schema update when running against dev db
    • fixing ejb reference
    • add support for specifying storage installer props file in rhqctl.properties
    • updating container ant script to generate rhq-storage.properties
    • make sure test cluster is shut down when there are errors prior to any tests running
    • fix regression in ClusterInitService.waitForClusterToStart
    • documenting the reason for ignoring exceptions on jmx query
    • we now need core/domain dep for C* dbreset
    • use separate executor and stream handler when running rhq-installer.sh --test
    • fix credentials for creating schema
    • move cluster shutdown into catch block
    • adding error logging and removing transaction support for adding numeric metrics
    • Merge branch 'master' into feature/cassandra-backend
    • updating pom version for metrics simulator
    • another pom version that i forgot to update in last merge
    • making scanForStorageNodes a @PostConstruct method and fixing ClassCastException
    • enable thrift for tests and remove unnecessary column in metrics_index table
    • add support for completely skipping cluster deployment to avoid the initialization overhead
    • major refactoring to prepare statements at start up and stop using batch statements
    • update baseline manager to obtain already initialized MetricsDAO object
    • moving test specific queries into test code
    • removing extra calls to ListPagedResult.iterator()
    • upgrade to C* 1.2.4 and upgrade driver to 1.0.0
    • update the branch we are building against
    • use EntityManager.merge instead of persist to handle both new and existing OOBs
    • adding debug logging
    • BZ 958087 ] shut down cluster object so agent can shutdown gracefully
    • adding rhq.cassandra.start_rpc property for token replacement in cassandra.yaml
    • updating schema for metrics_index table so that rows are not reused
    • need to pass timestamp to calculateAggregates now
    • fix setter method for startRpc deployment property
    • Merge branch 'master' into feature/cassandra-backend
    • changing transaction demarcation for basline and OOB calculations
    • initial implementation for using the CQL driver's async API for inserting raw metrics
    • changing transaction demarcation for basline and OOB calculations
    • updating logging
    • adding test to generate timeout exception in driver due to large number of inserts
    • log total insertion time
    • throttle async inserts with a semaphore
    • remove extra call to updateMetricsIndex and break up test into 2 large inserts
    • removing sync version of MetricsServer.updateIndex
    • Merge branch 'async-cql' into feature/cassandra-backend
    • remove class that was just for prototyping
    • Apply schema changes based based on cluster size
    • handle race condition during raw data inserts
    • re-applying test failure fixes that were lost during last rebase
    • use out of box default of 128 for native_transport_max_threads
    • do not update the system_auth keyspace replication_factor during schema creation
    • try to deal with async raw inserts
    • add log4j depenedency needed in order to build datastax driver module
    • set hostname of dev-container storage node to localhost by default
    • revert changes to ProcessExecutor.startProgram since they break api compatibility
    • disable C* schema updates by default
    • need to use reduced thread stack size for openjdk 7 on i686 arch
    • remove unused properties
    • need to used reduced heap sizes for C* on master jenkins jobs
    • refactor query to find schedules without baselines
    • removing unused method that has been replaced with more granular transaction demarcation
    • adding cleanup operation and some logging
    • inital commit for KeyspaceService
    • more refactoring to get the correct logic in KeyspaceService
    • Initial commit for StorageNodeComponent and new addNodeMaintenance operation
    • fix typo in ks name and add initial support for params in addNodeMaintenance
    • initial commit for resource operation to update seeds property in cassandra.yaml
    • adding support to update seeds property in addNodeMaintenance resource operation
    • fix parsing/writing the cassandra yaml config document
    • adding error handling to updateSeedsList operation
    • do not update the system schema directly
    • fix logging for datastax driver and for rhqctl
    • make sure seed nodes are accounted for and persisted at start up
    • need to initialize storage client before comm services
    • have to user cassandra super user to set up C* schema
    • add shutdown operation for storage client subsystem
    • remove references to log4j dep that is no longer packaged with datastax module
    • need to disable compression by default for now until it is made configurable
    • wrong option value was queried for compression
    • remove log4j reference from datastax driver module
    • disable client compression by default
    • adding initial support for upgrades
    • initial support for updating seeds list on upgrade
    • minor refactoring to separate install vs upgrade logic
    • log execution times for aggregation and add test for async raw data query
    • add support for aggregating data that was collected prior to server start up
    • only check if the full duration of the time slice has completed once
    • do not assume default jmx port when verifying node is up
    • adding support for deploying multiple storage nodes with dev-container
    • fix tests that were failing because of the storage client failing to initialize properly
    • fix bug for parsing jmx port and only set rhq.cassandra.seeds with local node info
    • provide defaults for heap and heap new size.
    • replacing CassandraClusterHeartBeatJob with StorageClusterHeartBeatJob
    • increase default timeouts for read, write, and range requests
    • fix integration test
    • temporarily disable test class while debugging jenkins issues
    • adding DataPurgeJobTest back for inclusion in jenkins build
    • do schema check via jmx instead of over cql
    • drop default cassandra user and perform schema changes with rhqadmin user
    • default to the IP address instead of hostname for listen_address property
    • use rhqadmin/rhqadmin as default username/password for storage node plugin
    • do not drop cassandra user. instead take away super user status and assign a random password.
  • Libor Zoubek <lzoubek@jezzovo.net> (4):
    • ircBot: improved recognition of commands
    • rhq-Bot: only devs can send private echo command to bot
    • ircBot: improved recognition of commands
    • rhq-Bot: only devs can send private echo command to bot
  • Libor Zoubek <lzoubek@redhat.com> (1):
    • BZ 968269 ] Make endpoints for pushing metric data more consistent
  • Lukas Krejci <lkrejci@redhat.com> (20):
    • Fix classloading issues when defining objects in the std script context.
    • BZ 928971 ] - ClassPoolFactory is now resilient against context classloader changes
    • Arq support for cleaning PC's data dir on shutdown.
    • @ClearPersistedData annotation handling in Arq plugin container tests.
    • BZ 928971 ] - Return the same class pool for the same context classloader so that the class cache of the class pool is consistent with what has been defined using given classloader. This should prevent the possible linkage errors when trying to define a single class multiple times in the same classloader which could have happened before this change.
    • BZ 928971 ] - Set the correct classloader for the dynamically created resource proxy classes.
    • Get rid of the dependency on the JMX plugin in the Arquillian container tests.
    • BZ 620603 ] - Don't use JOIN FETCH in criteria queries with limits.
    • BZ 620603 ] - Re-enabling the limited JOIN FETCH reporting with a more informative message, asking the users to co-operate in identifying the individual uses of the problematic JOIN FETCH queries that we potentially missed.
    • BZ 965833 - Potential bug in join/fetch/avoidance code] It is now possible to lazily fetch fields defined in super classes.
    • BZ 967622 - Server CLI script can't require modules from rhq://downloads]
    • BZ 965833 - Potential bug in join/fetch/avoidance code] It is now possible to lazily fetch fields defined in super classes.
    • Improving the explanation of rhq.rootDir. Adding a new rhq.dev.data.dir property in the server's pom.xml.
    • Trivial - improving the javadoc slightly.
    • BZ 969584 - HTTP Basic auth for bundle URLs] It is now possible to optionally specify the user name and password for HTTP URLs protected by basic authorization.
    • Make the size of the wizard dialogs overridable.
    • Adding API check exceptions for the newly added remote API methods.
    • BZ 969584 ] - Removing the confusing "Optional" hint in the field. There is a tooltip providing more detailed information on the fields.
    • Trivial - error reporting.
    • Removing the --agent option from "rhqctl console" altogether.
  • Mike Thompson <mithomps@redhat.com> (100):
    • BZ 924691 ] New graphs inconsistent coloring. Made hover label colors match the new actual chart colors.
    • BZ 921022 ] - Small Width Dashboard charts overflow.
    • BZ 921022 ] - Small Width Dashboard charts. Create small dashboard charts with same 8 scale as full dashboard charts just truncated to 20 bars instead of 60.
    • For Group Availability graph change WARN status to MIXED to better convey the meaning of the status.
    • BZ 948080 ] Add AutoRefresh capability to graphs other than portlet graphs. Refactored Auto refresh capability to be more generic so that it could be used for more than just portlets.
    • BZ 924942 ] - Add portlet key to New charts on dashboard so the same resource, metric can be displayed on multiple dashboards.
    • Fix NPE condition when autorefreshing a dashboard portlet graph before it has been configured (it has been added but not configured).
    • Fix issue with Resource Group Charts and refactor class hierarchy to make it flatter by eliminating AbstractMetricD3GraphView and ResourceGroupMetricD3GraphView.
    • Rename ResourceMetricD3Graph to MetricD3Graph to better reflect its new, more general, responsibilities.
    • BZ 949103 ] - On inventory detail summary metrics charts N+2 clicks of Set button stops firing.
    • BZ 872352 ] - Enable extended console methods not supported by IE from throwing undefined errors.
    • Trivial - add Log.error for codesplit async download failure so it is visible in browser console. Useful for after the fact diagnosis.
    • Trivial - remove client side Batik SVG transcoder dependency.
    • BZ 951196 ] - IE9 d3 charts X-axis spacing not very legible. Change the ticks to 8 ticks instead of 12 for IE9+ only.
    • BZ 958938 ] - Dashboard graphs need refresh when switching View/Edit portlet modes.
    • BZ 958754 ] - d3 multi-line chart: time units should have minutes when appropriate. Added the appropriate d3 time format string and updated the nvd3 chart library neither of which fixed the issue. It appears to be an error in nvd3 chart library.
    • Trivial: add *.DS_Store files to .gitignore
    • BZ 960646 ] - RFE Redesigned Availability Charts.
    • BZ 960646 ] - RFE Redesigned Availability Chart - Add disabled state cross hatch SVG pattern. Pixel tweeks.
    • BZ 960521 ] Globally uncaught exception on clicking on resource name link on IE9.
    • Make x-axis more readable on Safari, IE. These browsers even though they are using the same SVG fonts, render wider fonts showing up as labels running into one another. Reduced the clutter on the x-axis for increased readability.
    • BZ 960646 ] - RFE Redesigned Availability Chart - i18n.
    • Add new jQuery stuff.
    • BZ 960646 ] - UXD updates to new Availability Charts.
    • BZ 960646 ] - Trivial: strengthen opacity on new Availability chart.
    • Remove jquery ui and replace with bootstrap.
    • Add AvailabilityGraphType Generics.
    • Change charttype package to graph and also graphtype sub package.
    • Refresh charts using new javascript date slider.
    • Fix some of button behavior for d3 slider implementation.
    • i18n the Button bar.
    • Merge branch 'master' into mtho11/dual-slider-chart-control
    • Fix various behaviours between d3 slider and graphs.
    • Look and Feel Tweaks.
    • Replace brushable availability graph with pure date slider - DateSliderGraphType
    • Replace brushable availability graph with pure date slider - DateSliderGraphType
    • Pull d3 Date slider and place into actual graph instead. Decision was made in UXD meeting today.
    • Graph d3 brush slider is now on Graph and date button bar resembles that of logstash (logstalgia) style operation.
    • Bug fixes on d3 graph brushes.
    • d3 Brush handle modification to represent a physical handle.
    • Fix chart size on detail summary graphs as new chart is a bit larger with new date nav controls.
    • Update the d3 javascript libraries to 3.1.10.
    • BZ 958754 ] Time units should have proper units for multi-line graphs. Updated the nvd3 javascript libraries to 0.9.
    • Merge cassandra master to slider graph controls. Fix merge conflict in Messages_de.properties.
    • Fix merge conflict in Messages_de.properties.
    • Merge branch 'master' into mtho11/dual-slider-chart-control
    • First iteration consolidated metrics tab with sparklines and expandable graphs.
    • Second iteration at consolidated metrics tab with sparklines and expandable graphs.
    • Get rid of d3 slider (brush) graphing classes no longer used.
    • Second iteration at consolidated metrics tab with sparklines and expandable graphs.
    • Deprecate LiveGraphD3View.
    • Disable d3 sliding brushes until more solid.
    • Get availability working in new Consolidated metrics view.
    • Saving date ranges for graphs to user preferences.
    • BZ 971648 ] - Min Values in Graph Legend use NaN values to determine mean value.
    • BZ 971648 ] - Min Values in Graph Legend use NaN values to determine mean value. Bug fix for regression introduced with prior fix.
    • Reverting javascript date controls to smartgwt date controls.
    • Merge branch 'master' into mtho11/dual-slider-chart-control
    • Remove live graph D3 chart.
    • Add MeaseaurementTableView back in and remove bootstrap js classes.
    • Add smartgwt buttonbar for date range selection.
    • Save new date selection range to user preferences.
    • Save new date selection range to max 6 months from 1 year value as it is too large to save to DB (in ms)
    • Documentation and cleanup of code for new button bar date range editor.
    • Remove d3 consolidated graph view.
    • Make date range text change as the slider over the graph changes.
    • Persist the slider changes to user prefs when the user is finished dragging range across graph.
    • Add ability to call redrawGraphs from the javascript chart so it can refresh once we are done dragging a new time range.
    • Make chart refresh work with chart sliders and persistent values set by sliders.
    • Eliminate the default chart slider covering the whole graph when first brought up. Only after you click and start dragging will the sliders appear.
    • Get rid of handles on the resize path and allow hover layer to float above so that hovers can now be used in conjunction with chart sliders.
    • Make hover values for individual points (non-aggregate bars) different and not show duration, min, max, avg because it is not applicable.
    • Fix date range label incorrect bug introduced by removing fully opened brushes at graph start.
    • Make the date range text box larger and dynamically calculate the what the range (difference) is. For instance '2 hours'.
    • Show chart values for single points: Value: 7.99.
    • Fix double refresh issue and make refresh more robust by delaying the redraw a bit so that the code can keep up with the screen redraw (like we have done elsehwere).
    • Merge from master 6/11/2013
    • Sometimes the sparklines graphs are not rendered they show up as '...' because the jQuery method invoking the sparkline can't find the css class in the DOM because it hasn't been fully rendered yet even though it has been added to the DOM tree. This is more likely to occur on slow machines.
    • Add Custom... Date Range to new dual slider date bar.
    • i18n ButtonBar Date/Time Range editor.
    • Fix time portion of Time range editor.
    • BZ 951929 ] - Resource Metric Graph portlet doesn't refresh. Also applies to group metric graph portlet.
    • Merge branch 'master' into mtho11/dual-slider-chart-control
    • Undo changes for new screen that are not ready to be committed yet. Revert to working version.
    • Fix group availability graph vertical spacing issue.
    • Remove unused class.
    • Update the composite multi-line graph to use the new date range navigation.
    • Button alignment with form and form height change.
    • Remove bootstrap css and js as it is not needed now for dual slider chart control.
    • Eliminate Charting brush resize 'handles' code as that has been removed.
    • Remove the d3GraphView.addSetButtonClickHandler() as this is now done with new ButtonBarDateTimeEditor.
    • Remove old style Date Range control from MeasurementTableView.
    • Merge branch 'mtho11/d3-slider'
    • Fix issue with cut down version of Portal graph as it is loosing last value (off by one error).
    • On Graphs fix OOB metric lines to account for metrics of type units == 'percentage'
    • Keep Portal graphs from showing d3 brushes.
    • Add some space to Button toolbar for new Date/Time Range editor (per UXD request).
    • Use the minified versions of javascript libraries for d3 and nvd3.
    • i18n the date/time format for new button bar range editor. This is just the format for javascript so it is the date/time format for the range, shown while you are dragging the sliders. It is javascript (not smartGWT) that is handling the d3 brushes dragging. Updated the DE and CS date formats as well.
    • BZ 976926 ] On MultiResource Graphs allow deselection of Resources. Interactive graph features.
  • Rémy Maucherat <rmaucher@redhat.com> (2):
  • Simeon Pinder <spinder@redhat.com> (6):
    • BZ 964290 ] update LdapGroupManager to handle servers supporting server side query throttling.
    • Update some message translation indicators that were left out for LDAP.
    • BZ 964290 ] Add ui field to SystemProperties to specify minimum page sizes for LDAP group queries and their ui i18n components.
    • BZ 964290 ] Update query page size ordering and property details.
    • BZ 964290 ] disable query page control by default.
    • Fix bad LDAP localization indicator.
  • Stefan Negrea <snegrea@redhat.com> (113):
    • Update Eclipse project classpath entries for metrics related work.
    • Make the new server metrics and related modules maven and Eclipse friendly.
    • Merge branch 'master' into feature/cassandra-backend
    • Merge branch 'master' into feature/cassandra-backend
    • Move baseline calculation away from SQL db. The baselines will continue to be stored and retrieved from SQL but the data for the calculation will have to come from new metrics storage system. Also removed most of the references to the old way of calculating baselines.
    • Allow direct setting of schedule ids for balines.
    • Refactor tables (column family) names into an enum.
    • The statement binder needs explicit conversion to string otherwise enum value is used to create the statement.
    • Update metrics table enum to have default ttl for data.
    • Initial data migrator work, only aggregate tables included.
    • Simplify MetricsDAO interface to use only long type for expressing date&time. Actual date&time manipulations should be done in higher layers, the dao is only used fpr retrieving/updating data so simple time in millis is sufficient.
    • Add the basic algorithm for calculating baselines using Cassandra stored data.
    • Update metrics dao to use prepared statements as much as possible with the current driver.
    • Improve baseline calculation performance by quering for raw min and max. Also updated the average algorithm to use rolling calculations to prevent overflows (due to addition).
    • Update existing average calculations to use the rolling algorithm to prevent overflow. Move the initial implementation to the test class to check for rolling algorithm correctness.
    • Add unit tests for the metrics baseline calculation and arithmetic mean calculator.
    • Update data migrator to use interfaces to compact the code and also make the process configurable.
    • Update the Cassandra infrastructure code to avoid starting a new cluster if one is already running on the same ports.
    • An entire test can run in a millisecond. Update the code to avoid failures due to fast test runs.
    • Update metrics baseline calculation to perform data filtering on the client code.
    • Add raw metric migration code. Also refactor the code to attempt the migration operation a few times before failure.
    • Add missing precision to assert statement.
    • No need to have this interface deprecated since the data migrator needs to be a permanent utility.
    • Retrieve the full measurement schedule from the db to satisfy entity relationships. This will keep domain design and code clean.
    • Fix queries to match updated version of the measurement baseline entity.
    • Add two scripts created to help with the metrics data migration effort. One script generates random data representive of the estimation and the second parses the xml plugin descriptors to extract counts for metrics defined.
    • Add async methods to MetricsDAO for aggregate and raw metric insertion.
    • Update the data migrator to user async methods for insertion of metrics. Also separate the batched migration code from the all-at-once migration.
    • Add paged result sets mapped to domain objects to the metrics DAO interface. So far only raw metrics can be obtained via a paged set.
    • Update REST metric handler bean to get results for streaming raw metrics from Cassandra.
    • Move domain specific classes to metrics.domain subpackage.
    • Updates to the metrics module to support paging of results. Various other updates and refactorings to support this change.
    • Various other updates to support pagination of results from Cassandra.
    • Work around driver quirks and correct iterator usage. Also, add documentation to the paged results implementations.
    • Add a command line runner for the data migrator to assist with benchmarking and potential usage within the application. Various updates for the data migrator. Also, update the tool that generates random data to support sql configuration from command line options.
    • Add dependencies for data migration command line runner.
    • Make test to drain statement accumulator more reliable by draining every time the number of loaded statements is equal to or exceeds the limit.
    • Avoid using named queries to speedup the processing since raw data is sufficient. Run the migration process in a separate thread. Use the Cassandra batch directly instead of using an array to accumulate statements before sending them to Cassandra.
    • Add native delete queries instead of entity deletes. Also, add code to retry Cassandra inserts up to max number of allowed failures.
    • Update data purging to preserve data at least until the end of a table migration. There is no need to provide the option of purging data after each batch has been migrated.
    • A few more updates to the migration process. Mostly refactoring and API updates.
    • Various updates to data migration runner to support simple command line invocation.
    • Enable partial Oracle support. Add command line option documentation. And add better parsing for boolean options.
    • Add documentation and re-order methods to follow flow.
    • Small changes to the data migrator CLI to highlight the estimation and migration processes. Also, enabled Cassandra compression by default.
    • Update the RHQ plugin to be compatible with Cassandra 1.2.2. Drop support for 1.1.x.
    • Add manual mvn install to the snappy build script since it was not called during make process.
    • Add delete-only option to the migrator CLI.
    • Completely remove support for Cassandra 1.1.x because of changes in the plugin. Also, disable integration tests.
    • Renable dev-container deployments and move the executable plugin out of the dev container builds.
    • BZ 952781 ] Update the phase for packaging third-party libs to include them in the non dev-container builds. This problem was masked in the dev-container builds due to the way the jar is constructed and deployed.
    • Change the default database for the migrator to rhq, which is the official default RHQ database.
    • Add oracle spefic profile to bundle the ojdbc inside the data migrator.
    • Make Cassandra native port configurable and simplify the cassandra host options to take only hosts. The port is unique accross all the hosts and the thirft port is no longer need.
    • BZ 923458 ] The output of dependency:tree maven utility prior to any dependency version changes.
    • BZ 923458 ] Dependency tree output after all the changes that went in for the past week. Adjusted a few depedencies to reduce the diffs between pre&post updates.
    • Prevent the maven assembly plugin from overriding the main project artifact.
    • Fix to avoid skipping the first record of every migrated table.
    • Revamp the way to get information about the current node. The address, native port, and cluster name are now retrieved from the actual cassandra yaml file on discovery.
    • Integrate the experimental bulk export for Postgres into the data migrator code. Also, move all the migrator related code to a sub-package
    • Add configurable authenticator property to the plugin. The authenticator is detected on discovery and then used for CQL connections.
    • Update the plugin configuration propertiest to capture the file path for the yaml file. Also, remove and update other properties that were not used properly.
    • Add back all the transitive dependencies that were packaged by the plugin.
    • If the conf folder is absolute then do not base it on CWD. This can happen depending how Cassandra deamon gets started.
    • Add back missing snake yaml dependency.
    • BZ 958898 ] Avoid any possible overlap between two subsequent queries by specifying an order for the query.
    • BZ 958898 ] One more update to the queries to use exactly primary indexes created. Also increase query timeout considerably to avoid timeouts due to longer initial queries.
    • BZ 959591 ] Add back specific ant version based on local pom variables to force the download of the dependency even if not directly used in the pom. Without this, the pom would assume that the version referred here is the version from parent poms. This can cause the submodule artifact to miss ant binaries if not already in the local maven repository. (cherry picked from commit 2bb45d3c4c9d5edc71acf4380ce7a10a93937aae)
    • Update test to avoid disabled managers. This is especially important for the TagManager that can be disabled via build properties. (cherry picked from commit c9ccdbf61842dd115ce0d3067752e85dbb1db730)
    • Disable all the arquillian-integration submodules from building when maven.test.skip is true. Also, fix modules that depended on these modules to include them only if they are built. (cherry picked from commit 7e9678418a8ee2683aafb395c1cc25415ab375e3)
    • Update development version to 4.8.0-SNAPSHOT.
    • BZ 960044 ] Improve Cassandra discovery resilience to process discovery errors.
    • First attempt to populate storage node entities.
    • Default to normal operation mode for now to satisfy entity requirements.
    • Add more default timeout properties to the list of properties used to unitialize the entity manager.
    • Remove CassandraNode entity local to ccm modules in favor of the domain StorageNode entity.
    • Move the storage node discovery code from the session manager bean to the storage manager bean. Attempt to find the resources associated with the storage node seed list. Fill in the status of the storage node based on the resource information.
    • Add a streaming cursor data source to the list of available data sources and make it the default implementation. This will prevent requiring the database after the initial result, the rows will be just be streamed to the migrator in requested chunks.
    • Change the initialization annotation for storage node bean to comply with the stateless annotation.
    • Create the RHQ Storage Node plugin and update the Cassandra plugin to ignore resources these types of resources from discovery.
    • Update storage manager bean to link to RHQ storage node resources. Plus a few minor updates to the plugin to expose functionality to inheriting classes.
    • BZ 964229 ] Remove tokens trait since it gets large with virtual nodes enabled. The information is still avaiable via the tokens to endpoint map property.
    • BZ 958898 ] Update the estimation process to avoid count timeouts. Also, improved the estimation process by not counting the non-migration time (eg. query preparation) as part of the variable time estimation.
    • Clip unused line.
    • Update storage installer for Linux to start cassandra with a relative classpath. Before this change the Cassandra deamon was started with all the jar classpaths as absolute.
    • Add debug logging to server metrics classes. Also, a little cleanup for the metrics server class.
    • One more attempt to enforce session timeouts for Postgres due to driver ignoring hibernate imposed settings.
    • Update the metrics index table during the migration to insert metrics index. This will trigger aggregation on the recently migrated data if needed. The period to be added to aggreation is the from one aggregation slice before the last aggregation.
    • Split the big data migrator class and decouple functionality from the big main class.
    • Cassandra schema manager.
    • Move assembly plugin into a separate profile since it is not required for main artifact, only when testing with the main runner.
    • Update topology management to schema manager. No change to the external interface for the schema manager.
    • Change gc_grace_seconds updates to ALTER commands.
    • Add the topology change commands in the install and maven/test supporting modules.
    • Update the relationships between the storage nodes and corresponding managed resources when new storage nodes get discovered or removed from inventory.
    • Fixing a merge rebase import omission.
    • Update storage node bean to process nodes discovered but not set in the seeds property of the server.
    • Quartz job to start node maintenance. Also consolidate storage node functionality into the storage node bean.
    • Still use the classloader to get resource stream when code not packaged in a JAR file.
    • Create a jboss module for the data migrator to allow easy injection of jdbc drivers and command line execution.
    • Add a Windows script for the new data migration module.
    • A few fixes to the quartz job to use only SLSB calls.
    • The properties were not linked properly back to the configuration entity when using setProperties method.
    • Update storage node maintanance job to: 1) Run the topology change via the schema manager if necessary 2) Update the seed list 3) Retrieve live configuration from the storage service
    • Add Cassandra schema manager dependency to the list for server jar.
    • Remove data migrator from server metrics module. The code has been relocated in a separate module.
    • Revert "Mention the data migrator after install and provide it in the server downloads section."
    • Avoid packaging sigar.zip since the maven ear plugin does not know what to do with it and errors out.
    • Update the data migration upgrade option to use the newly created data-migration module and starting scripts.
    • Fix a test failure due to attempting to retrieve the simple name of an anonymous class.
    • Fix a test error where the alert was created referencing one resource but attempt to persist it referencing another resource.
    • Fix failing tests due to inconsistent setup and usage of entities.
    • Update release script to make correct use of boolean flags. Boolean flags are to be used without quotes in bash scripts, adding quotes will make bash interpert them as string variables.
  • Thomas Segismont <tsegismo@redhat.com> (34):
    • Add itests for AlertManager
    • Restore QUERY_NATIVE_TRUNCATE_SQL in Alert objects to preserve API
    • Stop compiling projects in verbose mode
    • AS7 plugin - use commons httpclient 4 as communication layer
    • as7plugin: increase delay so that SecurityModuleOptionsTest can discovery all resources
    • Fix Eclipse classpath file for httpclient 4
    • Attach animal sniffer plugin check goal to the verify phase of the build
    • Animal sniffer: verify is the phase to attach to, not the goal to execute
    • Animal sniffer: temporarily skip in portal war module
    • Fix failing itest AgentComm3Test
    • Fix itest failure in MeasurementScheduleManagerTest
    • ResourceMetadataManagerBeanTest: increase time to wait before types get removed (fails randomly on Jenkins)
    • Enable animal-sniffer-maven-plugin in portal-war
    • BZ 951738 - [hotfix] Use Commons HttpClient in ASConnection instead of JDK's HttpUrlConnection
    • Introduce ExistingDataSource abstraction
    • Add JPA bulk export source
    • BZ 911321 - [EAP 6.1] Reload operation fails in standalone mode
    • Add test case for Bug 911321 - [EAP 6.1] Reload operation fails in standalone mode
    • BZ 865629 - (PRODMGT-245) Add more metrics to AS7 plugin
    • BZ 962379 - [as7] Garbage Collector Resource misses 'collection-count' and 'collection-time' per minute metrics
    • BZ 962787 - GloballyUncaughtException when changing a parameter in Connection Settings tab of a compatible group after a plugin update
    • BZ 920214 - [as7] add async avail check support to AS7 plugin
    • BZ 885686 - Tomcat web application configuration form allows save without Config File required field
    • BZ 920214 - [as7] add async avail check support to AS7 plugin
    • Use a port from dynamic port range in integration tests
    • Bad port range in previous commit (PortNetServiceComponentTest)
    • BZ 865634 - Add more metrics to AS7 plugin
    • BZ 854773 - [as7] Unable to enable datasource on creation
    • BZ 956442 - Server script rhq-server.sh 'stop' command creates many shutdown threads that can exhaust user process limits
    • Fix DiscoveryBossBean: wrong injection type for storageNodeManager
    • BZ 970177 - Cannot browse the GUI while the plugin scan is running
    • BZ 924903 - JBoss AS JVM resource is not discovered after AS is restarted with jboss.platform.mbeanserver
    • BZ 975384 disable adding the sort by id.
    • BZ 975384 - Error displaying calltime data due to SQL error
Labels:
None
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.