Software Release Notice
for
CA Infrastructure Management
Release 2.3.4
Last revised 8/14/14
INSTALLATION INSTRUCTIONS
For detailed instructions related to product installation, see the CA Performance Center Installation Guide and the Data Aggregator Installation Guide.
2.3.4 provides fixes and enhancements to pre-existing functionality.
Please refer to the "DETAILS OF CONTENTS" section below for full
descriptions of these fixes and enhancements.
Symptom: In the Group Admin, when editing a rule where the 'Add' dropdown is set to 'Device Components', the accompanied description text - "A device component, like port, CPU and memory resource, etc." - is misleading because it implies that CPU and memory can be added to a group. This is not the case since CPU and memory are "unmanaged" components and to CA Performance Center they are just part of the device and not separate items. Resolution: Changed the description to read "A device component, like port, CBQoS and Response Path Test, etc." (2.3.4, 180588, 21302913-01)
Symptom: Top Flows by Interface graph is incorrectly named based on the data graphed. Resolution: Renamed the graph to be Top Flows by Volume to match title change made in Network Flow Analysis. (2.3.4, 225044, 21032114-01)
Symptom: There were some database changes to synchronize table sizes in the netqosportal and event management database not related to any defect. Due to an incorrect file specification, these were not applied in the 2.3 upgrade. Resolution: Added a new Event Management schema version script with the same changes. (2.3.4, 252142)
Symptom: When I go to add or edit a dashboard and select an NFA Aggregate or Virtual interface, I do not have access to any of the IP Flow views. Resolution: Modified the view definitions to allow them to be created for Virtual Interfaces or Aggregate Interfaces. (2.3.4, 254757, 21535713-01)
Symptom: LDAP Login fails with user-name that includes both lower and upper case letters. Resolution: LDAP authentication will convert the user-name to lower case before checking it against the LDAP server. (2.3.4, 262211, 21563728-01)
Symptom: When a data source is deleted, it can leave behind orphaned groups that no longer appear in the group tree. Resolution: Modified the upgrade script and the process of deleting a data source to create a "Data Source Orphan Groups" group under All Groups -> Inventory -> Data Sources and re-parent the orphans under this group. If there are no orphans, the group will not be created. These orphan groups can generally be deleted. It may take several minutes for the orphaned groups to appear in the group editor. (2.3.4, 269410)
Symptom: Getting the following error when running the Data Repository installation scripts: "Error: Unable to resolve network interfaces for 172.24.100.93." Resolution: The dr_validate.sh and dr_install.sh now explicitly set their LANG variable to English as they are not designed to be localized. (2.3.4, 275397, 21591429-01)
(2.3.4, 265082, 21571517-03)
Symptom: If there is a failed SQL query while running a discovery on the Data Aggregator, any in-progress discoveries may not timeout properly. Subsequent attempts to create new items upon discoveries will fail. Resolution: The discovery code was fixed to ensure that if a SQL exception was encountered, or connection to the particular DR node is lost during that stage, that we end up failing the discovery appropriately, so that subsequent attemps to re-run the discovery would be successful. (2.3.4, 278230)
Symptom: The custom reports from Network Flow Analysis are showing up under the inventory menu in CA Performance Center. Resolution: The custom reports from Network Flow Analysis are now suppressed when the Inventory list is displayed and global search is performed. (2.3.4, 280872, 21651109-01)
Symptom: The XML requests and responses from the Group Admin web service are inconsistent with the naming conventions from the user interface of CA Performance Center. Resolution: Changed the Group Admin REST endpoint to match the UI. "User Group" has been changed to "Custom Group" and "Site" has been changed to "Site Group". (2.3.4, 284806, 21665095-1)
Symptom: There are two Y axes for the same values in the Trend chart. Resolution: The ribdocs now contain the correct units for the baseline metrics instead of float. (2.3.4, 289491, 21693399-01;21743166-01)
Symptom: For some views tied to an Application Delivery Analysis data source, data column names are not translated into non-English languages properly. The column names could show as Unicode characters, such as \u8aac\u660e, instead of the actual foreign language characters. Resolution: Translation was modified to show the foreign language characters instead of Unicode characters. (2.3.4, 289912, 21708404-01)
Symptom: With the Calendar Heat view, any value that exceeds 100% is considered an outlier and is not properly processed when rendering the chart. Resolution: Data values greater than 100% are properly handled in the Calendar Heat chart when rendering the display. (2.3.4, 294506, 21718142-01,21730423-01,21638172-01)
(2.3.4, 283259, 21638172-01)
Symptom: The Detailed Poll Logging for Data Collectors does not contain sufficient information for debugging problems with polling of IPSLA tests. Resolution: The Detailed Poll Logging for Data Collectors was enhanced to record additional SNMP responses during polling of IPSLA tests in order to assist with debugging problems with polling of IPSLA tests. (2.3.4, 294775, 21668006-01)
Symptom: After creating an on demand Template with View Type "Chart per Item with Multiple Metrics" or "Chart per Metric with Multiple Items" with devices and "Metric Calculate Level" set to "By Component", you cannot change the Metric Calculate Level to by Device. Resolution: Modified the code to enable "by device" when appropriate. (2.3.4, 296636)
Symptom: LDAP authentication using GSSAPI could fail trying to find the configuration files for GSSAPI authentication if the service wrapper isn't configured correctly during install. Resolution: Installation of the Single Sign-On (SSO) component will update the service wrapper configuration to contain the path of SSO home so the GSSAPI configuration files will be loaded correctly. (2.3.4, 297658, 21733470-01)
(2.3.4, 312696, 21839298-01)
Symptom: When you view the Interface Capacity Watch Lists dashboard, the Top Interface Utilization Out Trend/Table and Top Interface Utilization In Trend/Table views do not load. It just stays in the "Loading data..." state. You may see the following exception in the PCService log: "java.lang.NullPointerException at com.ca.im.portal.plugins.rib.models.composite.RIBChartTableModel.createTrendChartModels(RIBChartTableModel.java:872)<br>at com.ca.im.portal.plugins.rib.models.composite.RIBChartTableModel.postUpdate(RIBChartTableModel.java:515)<br>at com.ca.im.portal.site.ModelMgr.updateModels(ModelMgr.java:712)<br>at com.ca.im.portal.site.ModelMgr.updateModels(ModelMgr.java:732)<br>at com.ca.im.portal.site.ModelMgr.updateModels(ModelMgr.java:732) ..." Resolution: Modified the code so that the view will display with no data since the group does not exist. The view(s) may be edited to reset them to factory defaults. If a user is deleting a group, they should review any dashboards they assigned that group to and reset them. Attempting to delete group references on dashboards when deleting a group (and its child groups) would introduce a potential performance issue. (2.3.4, 297687, 21737284-01)
Symptom: When customizing eHealth in a CA Performance Center view, if the context value is changed on the view setting form, an error 500 exception is generated. Resolution: Fixed context modifications error when attempting to configuration eHealth views with CA Performance Center. (2.3.4, 298179, 21729429-01)
Symptom: If a user attempts to create an on demand report template and they encounter the warning message about the metric filter attempt failing, and then try to use the filter box in the Metrics To Include panel for a search with over 50 results (by default), the prompt to perform the search may not be displayed properly and does not contain the "yes" and "No" buttons. Resolution: Modified the code to remove an old banner before displaying a new one. (2.3.4, 298234)
Symptom: After adding a custom view, views are missing from the default Dashboard Tab, and the Tab is no longer able to be edited. Resolution: A SQL script was developed that will fix the particular case of "Response Path Jitter" pages that were not fixed by the previous sp_fix_tabs.sql script. (2.3.4, 298863, 21744765-01;21792621-01;21793244-01)
Symptom: In some scenarios when an invalid custom metric family led to corruption in the TypeCatalog, unexpected exceptions were seen in the DA karaf.log when periodic change detection ran on a Metric Family. Resolution: Added checks for valid type catalog facet definitions such that improper custom metric family definitions would not prevent Change Detection from completing successfully on related metric families. (2.3.4, 298984, 21741364-1)
Symptom: When installed on a system that does not meet the minimum requirements for memory, CA Performance Center may become unresponsive and display timeout errors. Resolution: Modified the install to allocate a minimum of 1280m of system memory on under-sized systems or 1/6 of system memory for valid systems. The defaults were 768m or 1/8 of system memory. (2.3.4, 299344)
Symptom: Customer runs discover ....checks Discover History and sometimes sees the IP address displayed as an integer number, like 3414. Resolution: Correct IP or hostname will be displayed in the IP column for all devices. (2.3.4, 299394, 21771585-01)
(2.3.4, 299589, 21742426-01)
Symptom: When creating or editing an On Demand template and you search for a metric name, you must double click the "next" result icon to advance your search when your result lands on a metric family. Resolution: Modified the code advance the search with a single click. (2.3.4, 299562)
Symptom: Single Sign-On authentication using SAML2 fails if the IDP signs the response but not the assertion. Resolution: SAML2 authentication modified to validate the response's signature (if signed), and then validate the assertion's signature (if signed). (2.3.4, 300130, 21731200-01 )
Symptom: Running SsoConfig can generate a NullPointerException when using the Reset Value functionality. Resolution: Fixed the Reset Value functionality so it will not result in NullPointerExceptions. (2.3.4, 300298, 21737234-01)
Symptom: After network outage, devices not being polled and discovers timeout until the Data Collector is rebooted. Resolution: Increased the size of all polling statistics from 32 bits to 64 bits. (2.3.4, 300305, 21747772-01)
Symptom: When a user tries to view the Top Deviations From Normal (Interface) - Errors - Router view in CA Performance Center they get a RIB query syntax error. Resolution: Enabled the calculation of the deviation from baseline on the PctErrors metric family. (2.3.4, 302397, 21763227-01)
Symptom: When using a CA Mediation Manager-related context (BRS Network Stats) in the Dynamic Trend View, there is no Metric Family to choose from. This affects only CA Mediation Manager contexts.values. Using SNMP Context Type, they instead can see Metric Families to choose. Resolution: Modified the code to properly display all of the Metric Families in Dynamic Trend Views for CA Mediation Manager. (2.3.4, 302621, 21631202-01)
Symptom: Synchronization Failure of Data Aggregator with Performance Center after a discovery of new items. Resolution: Enforced a length limit of 255 characters for item description fields. (2.3.4, 302927, 21771721-01)
Symptom: Timezone 'America/Argentina/Salta' does not appear in any Timezone pulldowns in CA Performance Center. Resolution: Added the 'America/Argentina/Salta' timezone to CA Performance Center. (2.3.4, 303390, 21774746-01)
(2.3.4, 295683)
Symptom: Authentication using SAML2 fails with NullPointerException due to missing RelayState. Resolution: SAML2 authentication modified to print an error message in the Single Sign-On log when missing or malformed RelayState is encountered. Documentation updated on how to add RelayState to IDP, if needed. Enhanced debug of SAML2 authentication to try and diagnose future problems easier. (2.3.4, 304409, 21765068-01)
Symptom: Components of some metric families on some SNMPv2c devices are not discovered. Resolution: If GetBulk request fails, GetNext will be applied to re-read the table. (2.3.4, 304589, 21715582-1)
Symptom: CA Performance Center multiple item charts display duplicate business hour labels in the legend when different items have the same business hour type. Resolution: The business hours displayed in the chart legend only show unique business hours entry on a view. (2.3.4, 304912, 21749038-01)
Symptom: The CA Performance Center REST self-documentation does not provide a link to the groups REST endpoint. Resolution: The name attribute was missing on the description of the web service which caused it to not show up in the list. (2.3.4, 305440, 21665095-1)
Symptom: After an upgrade, the PC/conf/wrapper.conf file does not have the set value for the following setting, which limits the amount of memory the console process may use: # Maximum Java Heap Size (in MB)wrapper.java.maxmemory=. Resolution: Modified the install to properly restore the console wrapper.conf file as part of the install. If wrapper.conf.old exists and the wrapper.java.maxmemory line is empty, the value from wrapper.conf.old will be merged into wrapper.conf and wrapper.conf.old renamed to wrapper.conf.old.merged. After upgrade, system administrators should review the console wrapper.java.maxmemory setting to verify it is sufficient for their expected user and reporting loads. (2.3.4, 306668)
(2.3.4, 303910)
Symptom: Server names are truncated in PDF reports. Resolution: Modified the code so that long server names will now wrap instead of being truncated. (2.3.4, 307133, 21795891-01)
Symptom: After upgrading Performance Center that has been configured to use SSL, you cannot log in. Resolution: Modified the install to read jetty-ssl.xml to find the keystore name, back up that file, then restore that file after the install puts down new files. (2.3.4, 307142, 21790001-01)
Symptom: When creating a view as a copy of another view, and changing the group for this new view, if you modify the 'max per page' dropdown for the view it could revert back to the group from the copied view. Resolution: Modified the code so that the new view properly retains the newly selected group. (2.3.4, 308021, 21771480-01)
Symptom: After importing a custom dashboard using the dashboard web service, navigating to the page results in a HTTP 500 error and stack dump is displayed. Resolution: Added validation to strip invalid characters such as tab and end-of-line characters from the data in xml tags. Modified the unhandled exception to be a logged error with the stack dump in the logs so that the page will display. A view with invalid data may not display any data. (2.3.4, 308688)
Symptom: After rebooting the server your Data Aggregator runs on, the Data Aggregator is unable to start up successfully. Resolution: The ActiveMQ service is now setup properly for both fresh install and upgrade cases. (2.3.4, 308733, 21810484-01)
Symptom: After upgrading the Data Aggregator, depending on what order duplicate custom certifications in the deploy directory were read, the wrong certification file could be implemented. Resolution: When loading a xml file for a valid custom certification into the Data Aggregator (DA), the DA will look for any older xml for the same certification and rename them to a .xml.bak.DATE file so they are not loaded. (2.3.4, 308747, 21812864-01)
Symptom: When you select a link in the group "bread crumbs" on a dashboard, you are redirected to your default page instead of staying on the current page. Resolution: Changed the way the url is updated so parts do not get overwritten by accident. Also modified the way the time control navigates to the next page. (2.3.4, 309584, 21820408-1)
Symptom: The Discovery Profile dialog can display as a dialog with just a title and without any controls within it. Resolution: Modified the code to handle blank discovery profile names so that the dialog will display properly. (2.3.4, 309601, 21787459-01)
Symptom: On a pre-2.3.3 Performance Center installation where Anomaly Detector data source was added, upgrading to 2.3.3 could result in a MigrateDb error during upgrade. Resolution: Updated the Performance Center installer to not error if the Anomaly Detector data source type is already in the database during MigrateDb. (2.3.4, 309949, 21818714-01)
(2.3.4, 319101, 21882530-01)
Symptom: On-Demand report limits the list of components rendered based on the number of metrics picked. Resolution: Fixed On-Demand Report View Type "Chart Per Metric with Multiple Items" with Metric Calculate Level "by Component" to properly report limit of 15 components on a single chart. (2.3.4, 310035, 21817370-01)
Symptom: After an upgrade from an earlier release, views show no data and may show red error messages. You may see errors in the PCService log similar to: "ERROR | RIB/Engine | 2014-06-06 08:36:57,251 | com.ca.im.rib.engine.sources.SourceScanner | Unexpected error occurred while scanning for RIB sourcescom.ca.im.rib.repository.RIBException: Error binding RIB CA.IM.DA.MF.NormalizedAlcatelLucentVideoInterfaceInfo during stage 2Failed to bind type reference: CA.IM.DA.Types.FloatPerSecondAt RIB node type 'com.ca.im.rib.doc.TypeRefType', in RIB document 'CA.IM.DA.MF.NormalizedAlcatelLucentVideoInterfaceInfo', with properties: class: {class com.ca.im.rib.doc.TypeRefType} id: {CA.IM.DA.Types.FloatPerSecond} source: {null}<br>at com.ca.im.rib.repository.RIBDocument.bindReferences(RIBDocument.java:407)<br>at com.ca.im.rib.repository.RIBCollection.bindReferences(RIBCollection.java:233)" Resolution: Changed the version of the "types" document (CA.IM.DA.Types.xml) on Data Aggregator to version 1.0.5. Performance Center will download this new document and start using it, the next time it polls the data sources for the RIB Documents. (2.3.4, 311416, 21831017-01)
Symptom: When starting up Data Aggregator sometimes custom metric families are not loaded. Resolution: Added logging to keep track of what xml files Data Aggregator is attempting to load in order to better diagnose this issue (2.3.4, 311459, 21831758-01)
Symptom: Stored events in the Event Manager database are not pruned according to the specified retention time (which defaults to 30 days). You may also see errors similar to the following in the MySQL error log:2014-06-09 00:00:00 19583 [Note] Event Scheduler: [root@localhost].[em.e_rotate_event_partitions] event execution failed.2014-06-10 00:00:00 19583 [ERROR] Event Scheduler: [root@localhost][em.e_rotate_event_partitions] Data too long for column 'l_old_partitions' at row 217. Resolution: Modified the stored procedure to make some arrays used to build dynamic queries larger. Added some extra code to ensure clean up of very old records (older than the retention time). (2.3.4, 312977, 21798990)
Symptom: New threshold violations being generated in Data Aggregator were not seen in CA Performance Center and were not forwarded to Spectrum. Resolution: When merging any unresolved event items, log any that are duplicates, then merge what we can to allow processing and continue to pass along to the Spectrum application. (2.3.4, 313840, 21849730-01)
Symptom: Customer request to have Device Context support paging on multiple chart context tabs. Resolution: Added support on Device Context Page for paging on multiple chart context tabs (2.3.4, 313934, 21833773-01)
Symptom: During large volume of communication between the DA and the DC (such as large discoveries), we may encounter a failure to send some updates down from the DA to the DC. Resolution: No messages are lost even during large volume of communications between DA and DC. (2.3.4, 314284)
Symptom: After upgrade customizations are made to the Data Aggregator's bulk data export configuration, the configuration is reverted to default. Resolution: Changed the Data Aggregator installer to selectively restore the bulk data export configuration. Also changed the Data Aggregator and Data Collector installers to restore any custom configurations from the /opt/IMDataAggregator/apache-karaf-2.3.0/etc directory which are not part of the install process. (2.3.4, 315217, 21858464)
Symptom: After applying an Interface Filter to a monitoring profile to a device that has more than one Interface Monitoring Profile, polling stops for interfaces that do not match the filter, instead of being monitored by the other Monitoring Profile. Resolution: Fixed the code flaw. Polling will only stop when filtered by ALL monitoring profiles. (2.3.4, 315438)
Symptom: Whenever a user tries to create or modify event rules in CA Performance Center they get a popup message of "the violation threshold must be a number (e.g., 20, 20.5, .205)" Resolution: Added logging around when an error occurs trying to parse out the values of the event rule to be able to better diagnose these issues in the future. (2.3.4, 315518, 21835806-01)
Symptom: Scheduled discovery profiles times appear to shift when you make changes to them from CA Performance Center. Resolution: We now use the users timezone whenever making changes to scheduled discovery profiles instead of using the one which was set when originally creating the discovery profile. (2.3.4, 316294, 21830846-01)
Symptom: Heap Usage on the DC grows even with a small number of polled items ( 1000 ) in one day. Resolution: The memory growth in the DataProcessor triggered by attribute writes in the Data Collector was addressed. (2.3.4, 316421)
(2.3.4, 317378)
Symptom: The segmentation script looped showing only one projection being refreshed when it should have found two projections on an expanded cluster. Resolution: Updated the segmentation script to handle how projections are named on nodes that are part of the expansion of a cluster. (2.3.4, 316908, 21868752-1)
Symptom: After a failed upgrade due to running out of disk space on the CA Data Aggregator or CA Data Collector the services are unable to start due to missing configurations. Resolution: Modified the CA Data Aggregator and CA Data Collector installation process to determine if there is enough room to perform a backup and install and abort if there is not. (2.3.4, 317324)
Symptom: When you edit a tab on a server context page, the "Top Disk Utilization (Table)" view shown on the Disk Health tab is not available in the list of available views. Resolution: Added the XML tag to the view definition so that the page editor displays the view under a "Storage/Disk" category in the view list when editing a server context tab. (2.3.4, 317508)
Symptom: Heap Usage on the DC grows even with a small number of polled items ( 1000 ) in one day. Resolution: The memory growth in the DataProcessor triggered by attribute writes in the Data Collector was addressed. (2.3.4, 319582) (2.3.4, 316438, 21829648-01)
Symptom: New threshold violations generated on Data Aggregator were not seen in CA Performance Center and not forwarded to Spectrum. Resolution: When merging any unresolved event items, log any that are duplicates, then merge what we can to allow processing and continue to pass along to the Spectrum application. (2.3.4, 319730, 21849730-01)
Symptom: Under high event load and high user load, the following pages are slow to load: Events Display, Device Details Tab, Interface Details Tab Resolution: maxActive is set properly in fresh install of 2.3.4; upgrades will need to be adjusted manually. Fix may be manually applied to all previous versions. (2.3.4, 320972)