Fixed Issues
for
CA Infrastructure Management
Release 2.3.1
Last revised 10/1/13

INSTALLATION INSTRUCTIONS

For detailed instructions related to product installation, see CA Performance Center Installation Guide and the Data Aggregator Installation Guide.

Special Considerations

Release 2.3.1 is a complete release of CA Infrastructure Management. It can be used for a fresh install, or as an upgrade from versions 2.2.x through 2.3.0.

Improvements

CA Performance Center reporting views added "Total Rate" metric as an alternative to "Average Rate". Total Rate is computed as the sum of the volume related metric for all devices or interfaces, within a given group, divided by the time in seconds, per resolution cycle. The new total rate metric is available through custom reports.

Resolved Issues

Symptom: When creating a new dash board page that contains a Dynamic Trend view; when attempting to edit the Dynamic Trend view the metric family pull-down list is blank and a status error displays on the view setting's dialog.
Resolution: When creating a new dash board page that contains a Dynamic Trend view, a status error displays until the view is configured with items (group, device, interface).
Reproduction: Add a new dash board Drag the Dynamic Trend View on the pageEdit the Dynamic Trend View directly from the "Dashboard Builder" Until the items (group, device, interface) are configured for the view, the Metric Family drop-down is blank
(2.3.1, 119934)

Symptom: When a user is auto-logged out of their account, and auto-logged into another account, if that other account does not have access to the last group they were looking at, they are shown an error screen that says they do not have access.
Resolution: When a user is auto-logged out, the URL parameters are removed. This is better for security too. Then, when they are auto-logged in, they are taken to the default screen for that user.
Reproduction:
1. Log in to CA Performance Center and set up anonymous logins using the /opt/CA/PerformanceCenter/SsoConfig script.
    Set it so the anonymous user is the "user" account(2).
For help, see SSO administration documentation.
    While you are in that script, reduce the cookie timeout to 5 minutes to make it easier to duplicate
2. Go to the CA Performance Center UI as admin.
3. Create a new tenant.
   Note: I also changed the theme to use the CA-Gray theme so I could tell which account/tenant I was in.
4. Modify the "user" account so that its group permission is on a different branch than that of the new tenant, or not all.
5. Log out and then back in as the user of the new tenant. For example, if you named your tenant test1, log in as test1_user
6. Navigate to another dashboard, and verify that there is a GroupID parameter in the url.
7. Wait for the SSO cookie to time out. If you followed the previous instructions, it should only take 5 minutes, and the background should change from gray to blue.
I first got and error that I did not have permission for the current page, but after navigating to another dashboard got the error the customer is seeing.
The reason for this is that the group id is still in the url, but the anonymous user cannot see that group.
(2.3.1, 192395, 21330395-1)

Symptom: When Data Aggregator fails to shut down within the 5 minutes limit, Data Aggregator process will get killed. Logs will be moved to a backup folder. If that backup folder already exists, contents in the backup folder will be lost.
Resolution: The shutdown script has been modified to keep backup logs going back several shutdown attempts.
Reproduction: Upon shutdown command, if Data Aggregator fails to shut down within 5 minutes, logs were getting overwritten.
(2.3.1, 200600)

Symptom: When viewing the description for a rule, the database ID of a group is used instead of its actual name.
Resolution: The rule description now correctly displays the name of a group that was selected as part of its condition(s) and uses an internationalized name if available.
Reproduction:
1. Navigate to the Group Admin UI in CA Performance Center.
2. Create a new rule that has at least one condition that checks to see if an item is or is not a member of a group.
3. When viewing the description for the rule, the database ID of a group is used instead of its actual name.
(2.3.1, 213501, 21439397-1)

Symptom: After a password change when using external authentication, accessing the CA Performance Center debug pages may still use the old password, or may not allow access at all.
Resolution: Tightened up security around log in to debug pages. Added some logging to indicate when someone tries to access the pages using an externally authenticated account.
Reproduction: Set the default tenant account 'admin' to external authentication. Change the password. Access to the CA Performance Center debug pages may still use the old password, or not allow access at all.
(2.3.1, 217320, 21457788-02)

Symptom: After following the instructions to import CA Application Performance Management pages as part of adding a CA Application Performance Management data source, the dashboard pages do not appear.
Resolution: Interoperability with CA Application Performance Management 9.5 has been restored. Updated page definitions so they function with CA Performance Center Release 2.3 and added them to the pages installed by default. View suppression prevents customers without CA Application Performance Management from seeing the pages when turned on. The manual step to import the CA Application Performance Management views is no longer necessary.
Reproduction:
1. Follow instructions in CA Application Performance Management documentation to add CA Application Performance Management as a data source and import dashboards using the npcshell command.
2. Log in to CA Performance Center.
The CA Application Performance Management pages do not appear, even when dashboard suppression is turned off. They cannot be added as menu items in the menu administration screen.
(2.3.1, 219212, 21507821-01)

Symptom: The VMware Overview dashboard takes between 2 and 4 minutes to load. In general, Top N views with large amounts of historical data are slow to render.
Resolution: The SQL for Top N Views has been optimized to yield significantly faster RIB query execution times. The performance improvement of the SQL queries for Top N views - in isolation - varies from about 2 times to 10 times faster than pre-optimization.
In terms of dashboard loading times, the performance improvement may not be as apparent as it varies depending on the number of concurrent views on the dashboard and the background operations executing in the Data Repository at the time. However, you should notice an improvement for the first few Top N views that are rendered on any dashboard. This results in overall improvements in the dashboard loading times . The fewer the views on a dashboard the more apparent the overall improvement in Dashboard loading times.
Reproduction: Launch the VMware Overview dashboard from Dashboards > Infrastructure Health menu items.
(If this dashboard is not present, it may have to be added to the appropriate Menu item by navigating to Admin > Menu > Select "Infrastructure Health" > Edit. Then find VMware Overview and move it to the "Selected Dashboards".
(2.3.1, 219413, 21441481-1)

Symptom: When logged in as default administrator, the menu option on the tabs of a context page that allows you to edit the tab, and so forth, is not available. Also, the "Edit Context Pages" role right does not appear as a granted or selectable right in any of the predefined roles.
Resolution: Modified the script to remove a space before the first line of sql which adds this role right.
Reproduction: Not reproducible.
(2.3.1, 220934, 21473131-1)

Symptom: Enable the "My Custom Groups" feature for users. Users cannot create custom groups using the 'Existing' tab and selecting a group from under the 'All Groups' area. When trying this, the user receives an error that states: "The group reference was not saved. Only groups belonging to the same tenant can be added as a reference to this group."
Resolution: Checks added to detect the edge case after all other checks have been performed.
Reproduction:
1. For any user, even admin, enable the "My Custom Groups" feature.
2. Give the user access to "All Groups" for the remaining group permissions.
3. Log in as the user being used for testing and go to the My Settings->My Custom Groups area. If you try to create a copied group using the 'Existing' tab in the "Add Group" UI, and select a group under the users "My Custom Groups", no error is seen.
4. Try the same thing, but this time use groups that are created under the 'All Groups' area as the reference group.
I see the same error as the customer that states: "The group reference was not saved. Only groups belonging to the same tenant can be added as a reference to this group."
(2.3.1, 220978, 21480948-1)

Symptom: When you select the Custom IM Heat Chart, and set the zones so that the highest value is green, and the lowest value is red, the chart displays all items in red, regardless of the zone settings entered.
Resolution: Modified the code so that the heat chart displays data properly when setting up the zones to have good values as high, and bad values as a lower value.
Reproduction:
1. Go to the page builder.
2. Set the context to a server, router, or switch.
3. Select the Custom IM Heat Chart and drag it to the page builder display. Save the page.
4. Click the gear menu and select "Edit".
5. Select the Availability metric family, and choose "Availability - Average".
6. Set the zones as follows to match the Availability heat chart zones on the device context page Green: 100 Yellow: 93 Orange: 83 Red: 0. Click Save.
The chart displays with all red values, when a 100% available device should show green. The Heat Chart on the context page works as expected with these values.
(2.3.1, 220997, 21484411-01)

Symptom: If you perform a search in CA Performance Center, modify the column settings in one of the grids and then try to save it, you get an error "Error 500 Model Identifier must not be zero".
Resolution: Assigned a model id and other necessary identifiers to the grids on the search page so they can be uniquely identified and have their column settings saved.
Reproduction: In the menu in CA Performance Center, perform a search for something that should return results, such as searching for a discovered server by name. In one of the grids returned, hover over the right side of one of the column headers, a small arrow icon should appear. Left-click this menu, navigate to "Columns" and clear the selection for a column to hide it. Now hover over "Save column settings" and select one of the available choices to save the settings. It should save your selection like other grids.
(2.3.1, 221217, 21487202-02, 21484025)

Symptom: When using custom vendor certifications which polled for data that should be a negative number, the values being returned were positive numbers (as they were being converted to unsigned integer incorrectly) causing a discrepancy between what was being displayed in CA Performance Center, and what the actual values were according to the device.
Resolution: To resolve the bug, the code which parses the raw data coming back from the device was modified to accept negative and positive numbers.
Reproduction: To reproduce, you must set up a simulator or device which can send a negative number as a response during SNMP polling, and a custom vendor certificate which polls for that negative number. Then view the data that is coming in and check that it has been converted to a positive number and not left a negative number, and perform an snmp get on the device to confirm that it should be a negative number.
(2.3.1, 221843, 21471005-1 )

Symptom: In rare instances, a data source may stop synchronizing and the status will be "Unable to contact", even though it can be contacted outside or performance center. No errors appear in Device Manager Log. Selecting the data source from the list page and clicking "Test" fails, while selecting "Test" from the data source administration dialog succeeds.
Resolution: Modified the code to only update the rows in this table if the data source sends updated information, never to delete them as part of the normal data source poll cycle.
Reproduction: Not reproducible under normal operation. Manually deleting the rows for a data source from the data_source_interfaces table shows the after-effects.
(2.3.1, 221851)

Symptom: When you are working in the Monitoring Profiles section of Data Aggregator and right-click a monitoring profile, a tool tip may obscure the menu.
Resolution: Modified code to dismiss and temporarily disable tool tips until the menu is dismissed by clicking outside the menu, or a menu item is selected.
Reproduction: Navigate to the Data Aggregator Monitoring Profiles page. Hover over a monitoring profile. When the tool tip appears, right-click with the mouse to bring up the menu to perform an action such as editing, copying, or deleting a monitoring profile.
(2.3.1, 222060)

Symptom: When you have multiple copies of a dashboard and delete one, depending on the parent of that dashboard, it may not delete the one you told it to.
Resolution: Modified the code that generates the URL and the javascript that executes the delete method so that it points to the correct parent which we want to delete the dashboard from.
Reproduction:
1. Create a new dashboard under Infrastructure Health, and then go to Admin>Menus, and edit Operations Displays.
2. Under Operations Displays, add your new dashboard to the selected dashboards and click save.
3. Confirm that they both show up under the dashboards menu.
4. Go to the Operations Displays instance of the dashboard and delete it, then look under dashboards again and see that the Operations Displays instance of the dashboard is still there, however the Infrastructure Health instance has been deleted.
(2.3.1, 222101, 21488097-01)

Symptom: Order of Dashboards are out of order compared to the explicit order they are put in
Resolution: Confirmed changes made to menu reordering logic and updated rank of many out-of-the-box dashboards to help ensure placement.
Reproduction: Custom Menus (or even out of the box menus) and custom dashboards. If you go to Menus->Edit (select a menu to edit) and without changing anything, close (Save) it. Go back and view the Menu lists. The dashboards shuffles how they are listed.
(2.3.1, 222105, 21487951-01)

Symptom: When you perform a search after changing the group, the search results may revert to the previous group that was selected.
Resolution: Updated user session property that stores user's last selected group when selecting intermediate groups from bread crumbs.
Reproduction:
1. Open "Infrastructure Overview", or any other dashboard.
2. Select any group filter (other than "All Groups" and apply it to the dashboard)
3. Type search criteria in the "Search All" control.
4. Return filter to "All groups" by clicking on the "All Groups" section of the page context indicator.
5. Type different search criteria into the "Search All" control.
(2.3.1, 222199, 21492628-1)

Symptom: When trying to access the "Collections" group from the Group Admin UI, a pop-up dialog displays, indicating that an error has occurred.
Resolution: The server now properly inserts data into a temporary table rather than trying to insert into a temporary table with the name of "en-US" or whatever the user's locale is.
Reproduction: Attempt to navigate to the "Collections" group from the Group Admin UI.
(2.3.1, 222255, 21499753-1)
(2.3.1, 223110, 21497413-01)

Symptom: When you attempt to discover the metric families of a device, you see that the metric family supported, however no components are listed in CA Performance Center, even if they exist.
Resolution: Modified the polling logic to check for a "too big" error in the response and adjusted the size of the request based on that.
Reproduction: Set up an SNMP agent which has a low value for the maximum size of a response allowed, using a bulk get request, and begin polling that device using CA Performance Center. Poll for a metric family which has many components (for example Interface), and see that there are no components available. If you perform a walkthrough of the device using MIB browsing utility, you may find that there are components which were not discovered by CA Performance Center.
(2.3.1, 222658, 21492032-01)

Symptom: When trying to assign a large number of permission groups to a user using the RESTful User service, the response time is too long. Assigning approximately 2500 groups takes about 10 minutes.
Resolution: Implemented various performance optimizations implemented to improve the throughput when validating all the permission groups to be assigned to a user.
Reproduction: Use any RESTful client to perform a POST operation of a large number of groups to the following URL:
http://<capc_host>:8181/pc/center/webservice/user/userName/<specify_user_name>/groups.
The response time will be long for more than 1000 groups.
(2.3.1, 222677, 21487455-1)

Symptom: When Data Aggregator is at steady state (polling metrics, but undergoing minimal configuration changes) and high scale, beginning at approximately 30 minutes after each hour, the Vertica logs begin to show messages containing the text "WOS Full; failover to DIRECT". Beginning at approximately 30 minutes after each hour, Vertica begins to create many ROS containers, and some may be small.
Resolution: The baseline and roll-up data for metric families whose hourly data consumes >100MB will be written directly to ROS. This mitigates the hourly WOS overflow. Smaller metric families, all rate data, and data repository configuration changes will continue to use WOS.
Reproduction: Operate Data Aggregator at a medium to large scale configuration.
(2.3.1, 223975)

Symptom: Full garbage collection happens frequently
Resolution: Change from '-DisableExplicitGC' to '+DisableExplicitGC'
Reproduction: Occurs automatically as part of stale event state clean-up.
(2.3.1, 224095)

Symptom: Slow response of CA Performance Center. mySQL consumes 100% CPU and CA Performance Center hangs until it eventually returns to normal loads.
Resolution: Identified a set of slow queries and optimized the queries to perform better with a large scale database.
Reproduction:
1. Discover and manage a large number of items (1/2 million or more) and define large groups using group rules.
2. Monitor CPU utilization on the CA Performance Center system during periodic data source synchronization, enable and monitor the mysql slow query log watching for queries that exceed 10 seconds to execute
3. Monitor the mysql process list ('show processlist') during periods of high CPU utilization on the CA Performance Center system.
4. Preview and run group rules within the CA Performance Center UI for a large group.
5. Log in to the CA Performance Center system.
(2.3.1, 224316, 21478889-01)

Symptom: Run report on Cisco device that has both low-speed and high speed interfaces does not report correct value.
Resolution: Fixed the high speed interface vendor certification to complain to RFC 1573:
- For interfaces that operate at 20,000,000 (20 million) bits per second or less, 32-bit byte and packet counters must be used.
- For interfaces that operate faster than 20,000,000 bits/second, and slower than 650,000,000 bits/second, 32-bit packet counters must be used and 64-bit octet counters must be used.
- For interfaces that operate at 650,000,000 bits/second or faster, both 64-bit packet counters and 64-bit octet counters must be used.
Reproduction: Discover a Cisco device that has both high speed and low speed interfaces. High speed interface vendor certification will be discovered. The data in report is not correct.
(2.3.1, 224384, 21495029-01)

Symptom: When saving a user-created dash board page, the contents of the page look blank; and the configured views are at the very bottom of the page.
Resolution: When saving a user-created dash board page, the contents of the page look blank.
Reproduction:
1. Add a dash board.
2. pageDrag a view onto the pageSave the dash board page.
The page may appear blank to the naked eye, but at the very bottom of the page there is the view added to the dash board page.
(2.3.1, 224497)

Symptom: Device name does not get updated on CA Performance Center after the hostname of the device name has been changed.
Resolution: Implemented periodic host name re-evaluation (default every 12 hours) to update HostName and trigger item name re-evaluation based on name ranking.
Reproduction:
1. Discover a device with Hostname be the first one in the naming order. Note the device name on CA Performance Center is the hostname of the device.
2. Change the hostname of the device.
The device name on CA Performance Center does not get updated.
(2.3.1, 225659)

Symptom: When you search from the menu in CA Performance Center, the title in the search results only goes to 1000 when the results indicate there are more results. You are not limited to 1000 results.
Resolution: Modified setup of search results to use the application value "UniversalList.Limit", which defaults to 5000.
Reproduction: In the menu in CA Performance Center, enter a search that will return a large number of results.
(2.3.1, 226307, 21512655-01)

Symptom: When you go to the User List page and have a large number of user accounts configured, the display of the list page may be slow.
Resolution: Optimized the generation of the list and how it displays.
Reproduction: Configure a system with many user accounts (2500+). From the main menu, select "Admin" -> Users, the list may take a while to display. This may occur also if you are navigating from the Roles List page.
(2.3.1, 227748)

Symptom: The upgrade to CA Infrastructure Management 2.3 may fail with an error, "Performance Center upgrade was not successful".
Resolution: Modified problem command in script to add valid rows, even if two or more of the default menus were not found.
Reproduction:
1. Install any version of CA Infrastructure Management 2.2.
2. Log in and go to "Admin" -> "Menus"
3. Delete at least two default menus, such as "Infrastructure Health" and "Capacity Planning".
4. Upgrade to CA Infrastructure Management 2.3.
(2.3.1, 251631, 21528616)

Symptom: In CA Infrastructure Management 2.3.0, if a collection was configured to contain both devices and components (for example, interfaces), when the collection is synchronized from CA Performance Center to Data Aggregator, the components also get added to the memberships in the data source's database.
Resolution: For collections, ignore components when synchronizing to Data Aggregator.
Reproduction: In 2.3.0, on CA Performance Center, create a collection with filters that include some device components and some devices. After a while, check the item_relationship table in Vertica. The components contribute to the member count for the collection.
(2.3.1, 251916, )

Symptom: During upgrade of CA Performance Center (from 2.2.x), if mysql upgrade fails to take a backup of CA Performance database, it fails with an error "Automatic backup did not succeed.” After you complete the upgrade and perform a manual database backup, the following databases are missing: em and netqosportal .
Resolution:
- Improved the error checking to exit if there was a problem with running the database backup, leaving the data intact.
- You must perform a manual backup using the command, (see "Take Additional Steps for Database Upgrade" in the CA Performance Center Installation Guide), which uses the same command that the upgrade uses to back up the database.
Reproduction:
During an upgrade from 2.2.x to 2.3.x, which involves a MySql upgrade, the upgrade performs an automatic database backup, which the following two steps.
1. mysqldump of the database to generate databaseBackup.sql
2. sed command on the databaseBackup.sql
If mysqldump fails (due to mysql or mysqldump utilities not being found or a problem during backup), the upgrade continues instead of exiting.
The upgrade then removes the database and upgrades MySql to 5.6, after which it tries to restore the database but fails to find it and exits.
(2.3.1, 251927, 21508605-01)

Symptom: When restarting the Data Aggregator, it may fail to load the items. CA Performance Center may give the error "enum.datasourceerror.DS_PRODUCT_ID_CHANGED" and you may see the following exception in the DA karaf log: com.ca.im.item.db.DbException: Failed to load attribute data.
Resolution: Increased the loading timeout from ten minutes to five hours, longer than it will take to load the item repository.
Reproduction: This may be caused by an out-of-conformance Data Repository, causing the initial item loading to take longer than the original 10 minute limit on a system with large amounts of items.
(2.3.1, 254452, 21516581-01)

Symptom: When manually running rules for a report group in CA Performance Center, a user may receive an error message stating that a database table could not be found.
Resolution: The database table that is leveraged by the CA Performance Center rules engine is unique for a given thread. This makes sure that no race condition occurs if rules for a report group are executed concurrently.
Reproduction: There are two known scenarios that cause this issue:
- Multiple users try to manually run or preview rules for the same report group at the same time.
- A single user manually runs rules on a report group and then makes and saves changes to that same report group (such as name or description).
CA Performance Center requests a resync to notify data sources of these changes. After a resync , the rules portion of sync runs and can overlap with the manual execution of the rules invoked by the user.
(2.3.1, 255290, )

Symptom: When the filter rules on a collection are modified, such that some of its members must be removed, the removals fail to occur on Data Aggregator.
Resolution: Fixed the logic so deletion occurs.
Reproduction: For a collection, modify its filter rules so some devices in it are filtered out. Check the item_relationship table in Vertica database and the member count for this collection does not change.
(2.3.1, 255291, )

Symptom: The memory consumption of Data Collectors grows over time and may eventually lead to out-of-memory error.
Resolution: The default log sizes have been reduced. They are also configurable through com.ca.im.dm.core.collector.cfg. Contact CA Support if there is a need to change them.
Reproduction: Data Collector, that polls many devices or items will, over time, consume memory.
(2.3.1, 256238, 21539084-1)

Symptom: CA Performance Center may stop synchronizing with Data Aggregator and the following error may appear in the Device Manager log: “An error occurred during a sync request with data source Data Aggregator: additional info: enum.datasourceerror.DS_PRODUCT_ID_CHANGED.”
Resolution: The ability to rename the Default Tenant has been disabled. A future release will include a solution to re-enable this functionality.
Reproduction:
1. Log in to CA Performance Center as the default administrator.
2. Navigate to "Admin" -> "Tenants"
3. Select "Default Tenant", then click "Edit".
4. Change the name from "Default Tenant" to another value and click "Save"
5. Wait a few minutes for CA Performance Center to synchronize the change to Data Aggregator.
6. Stop, and restart Data Aggregator. When it comes back up, CA Performance Center may stop synchronizing and log an error message.
(2.3.1, 256592, )

Symptom: The Data Aggregator restart time increases dramatically.
CA Performance Center "Manage Data Sources" page shows communication or version failure messages when clicking Log for the Data Aggregator data source.
The karaf.log file on Data Aggregator shows error messages similar to the following text, starting at approximately 2:00 AM (default, but configurable time):
StatementCallback; uncategorized SQLException for SQL [SELECT DROP_PARTITION...A Moveout operation is already in progress on projection...
Resolution: The Data Aggregator partition maintenance process no longer creates duplicates of the records that are still current. This reduces the complexity of the query for Vertica.
Additionally, Data Aggregator now requests Vertica to ignore data in memory when dropping partitions so the partition will be dropped when Data Aggregator requests it.
The data in the partitions being dropped has been in Vertica for days and will no longer be in memory, so no loss of current data occurs when the partitions are dropped.
Reproduction: The reproduction scenario is based on timing. Vertica must be in the process of moving data from memory to disk for the specific table, while Data Aggregator is trying to drop a partition on that same table. Data Aggregator attempts to drop a partition shortly after 2:00 AM on the first night after upgrade to 2.3.0, then, subsequently on the eighth night and each successive night after upgrade to 2.3.0.
(2.3.1, 256993, )

Symptom: In inventory discovery, for each ping-responded IP, if there is no existing device item found at device item's primary address, a new device item will be created to perform device discovery. The device discovery collects all necessary information and runs the reconciliation process to see if the device was already discovered at a different IP or not. If yes, the newly created device item will be deleted.

For devices that have many secondary IPs, this process is very expensive. Re-discovering a large network can seriously impact CA Infrastructure Management performance.
Resolution: For each ping-responded IP, if it can be found in any existing device item's IP address table, the device at that IP is marked as already discovered. This prevents creating a new item and running the device discovery and de-duplication for the item. It can be configured; default is disabled.

To enable this evaluation feature:
1. Go to $DA_INSTALL/apache-karaf-2.3.0/etc directory.
2. Create file com.ca.im.tm.inventory.discovery.impl.ProfileDiscovery.cfg with this line in it:
lazyDeviceEvaluation=true
3. To disable the lazy evaluation function, change the line to
lazyDeviceEvaluation=false
Reproduction: Re-discover large IP network. Data Aggregator may become unresponsive. Data Collectors may lose connection to Data Aggregator.
(2.3.1, 257518, )

Symptom: Data Collector may disconnect from Data Aggregator. The ability to rename the Default Domain was disabled.
Resolution: The ability to rename the Default Domain was disabled. A future release will include a solution to re-enable this functionality.
Reproduction:
1. Log in to CA Performance Center as the default administrator.
2. Navigate to "Admin" -> "IP Domains".
3. Select "Default Domain", click "Edit".
4. Change the name from "Default Domain" to another value and click "Save"
5. Wait a few minutes for CA Performance Center to synchronize the changes over to Data Aggregator.
6. Stop, and restart Data Aggregator. When it comes back up, Data Collector may disconnect from Data Aggregator.
(2.3.1, 257798, )

OVERVIEW for 2.3(Back to DESCRIPTION)

2.3 provides fixes and enhancements to pre-existing functionality.
Please refer to the "DETAILS OF 2.3 CONTENTS" section below for full descriptions of these fixes and enhancements.

DETAILS OF CONTENTS

Resolved Issues (Back to OVERVIEW for 2.3)

Symptom: When polling CBQoS on Cisco Nexus 7000 devices, the device soon stops responding to SNMP requests.
Resolution: Modified CBQoS vendor certifications to not poll interfaces that are Administratively Down, or interfaces of type Loopback and Other.
Reproduction: Not applicable.
(2.3, 173446, 21281117-01)

Symptom: Installer failed to report failure during installation.
Resolution: Fixed the installation.
Reproduction: Not applicable.
(2.3, 174460, 21280323)
(2.3, 92862)

Symptom: Unable to navigate to the original/real group from a reference group in the Group admin UI of CA Performance Center.
Resolution: The link used to navigate to the original/real group in the Group admin UI of CA Performance Center now works.
Reproduction:
1) Navigate to the Group admin UI in CA Performance Center.
2) Create a new user or site group.
3) Make a copy/reference to that newly created group.
4) Navigate to the copy/reference of the newly created group.
5) Click the link that is supposed to take you to the original/real group that is being referenced. Nothing happens.
(2.3, 176108, 21287555-01)
(2.3, 187947, 21331299-1)

Symptom: The user cannot remove a reference group from the "All Groups" and "Service Provider Global Groups" groups.
Resolution: The Group Admin UI and corresponding back-end services now allow reference groups to be removed from either the "All Groups" and "Service Provider Global Groups" groups.
Reproduction:
1) Create a new group under either the "All Groups" or "Service Provider Global Groups" group and select the "Existing" tab and choose an existing group to reference.
2) After the reference has been created, navigate to original group that is being referenced.
3) Click the "Remove References" tab to remove the newly created reference. The user will encounter a misleading pop-up dialog stating "CA Performance Center is currently synchronizing, try again later" and the reference group still exists.
(2.3, 176132, 21287555-01)
(2.3, 181226, 21278492-1)

Symptom: Notifications sent for Event Manager events may contain a URL from which you cannot access CA Performance Center depending on the configuration of the CA Performance Center web server.
Resolution: Code changes to help ensure database field that contains appropriate IP address is included in events, allowing launch backs to function correctly.
Reproduction: Configure CA Performance Center with two active NICs, one having a public management IP address, and the other a private IP address.
(2.3, 177142, 21269045)

Symptom: When you select a dashboard, you must wait for the entire dashboard to load before changing the group. Otherwise the views start to get jumbled up between the initial group and the latter group. Some of the views will be set to the first group and some of the views will be set to the group selected while the dashboard was loading.
Resolution: When the group context is changed, it signals a page refresh and updates the GroupID URL parameter (if available) to the updated group's ID to help ensure that the proper group context is used. On page load, views will use the latest group context. If the context is switched again, the page will once again refresh.
Reproduction: Select a dashboard and change the group before the dashboard finishes loading.
(2.3, 178322, 21295128-1)

Symptom: In CA Performance Center, the browser's back button could go too far back and remove a selected group.
Resolution: The URL now encodes the group ID when appropriate, via the GroupID parameter.
Reproduction: When you choose a Dashboard and change the group, then select to drill into a table item for that group if you then select the browser's back button, it backs out to the Dashboard but does not keep the group that you had changed to.
(2.3, 178331, 21295191-1)

Symptom: Dashboard that is created in "My Dashboard" should not available for selection in Menu Editor.
Resolution: Dashboard that is created in "My Dashboard" will not be available for selection when editing menu
Reproduction:
1) Create a new dashboard and save for all tenant users saved under 'My Dashboards'.
2) Log in as Administrator, and navigate to 'admin->menus'.
3) Select one of the out of the box menus to edit.
4) The newly created dashboard can be selected.
5) Save the menu.
6) The newly created dashboard does not show up in the menu.
(2.3, 180308, 21270074-1)

Symptom: Device Page shows that the device is not a member of any group even though the device is in the "All Group".
Resolution: Product fixed to show the correct group(s) to which the device belongs.
Reproduction:
1) Create a new group under All Groups named Empty.
2) Create a new user account, and add both Empty and All Groups to the permission group list.
3) Log in as the user, and select the All Groups entry from the dashboard group navigation area.
4) Drill into a device, and the group membership area shows no groups.
(2.3, 180342, 21303259-1)

Symptom: Graphs for "Bits In" / "Bits Out" on interfaces have gaps in them due to 32-bit counters rolling over every 3 poll cycles or so.
Resolution: Two changes were made to resolve this issue:
1) Added in the ability to compute deltas even when 32-bit and 64-bit counter wraps. This option is off by default, and must be turned on for each data collector. Create a file named "com.ca.im.dm.snmp.rdp.impl.SnmpDeltaController.cfg" and place it in the Data Collector's $KARAF_HOME/etc directory. Adding the following line should dynamically change the behavior of 32-bit and 64-bit counter wraps: "showGapsOnCounterRollover=false"
2) The maximum value of ifInOctets and ifOutOctets deltas was fixed at approximately 784 million. Any interface being polled with low-speed counters shouldn't be greater than 20 Mbps. Anything higher than that should be polled with the high-speed ifXTableMib counters. If the delta value of these counters is greater than 784 million, then a gap will still be shown in the Bits In, Bits Out graphs respectively.
Reproduction
(2.3, 180809, 21288167-1)

Symptom: User is cannot remove a reference group in CA Performance Center and the following message is displayed: "CA Performance Center is currently synchronizing, try again later".
Resolution: Fix to disable the option to remove child of reference group.
Reproduction: Try to remove a reference group in CA Performance Center.
(2.3, 181226, 21278492-1)

Symptom: CA Performance Center slow performance when CA Spectrum 9.2.2 is integrated as a data source.
Resolution: Enabled periodic analysis of the Event Manager schema in the database and improved several of the queries used by the CA Spectrum integration to perform better. Also, added a two minute timeout so that any long running queries will be killed and an error returned rather than starving out the rest of the system.
Reproduction: Integrate CA Spectrum with a substantial number of devices as a data source and export multiple landscapes.
(2.3, 182795, 21289606)
(2.3, 181647, 21289606)

Symptom: After user edits or renames a tab from the CA eHealth Router page, user cannot open the router context page with the following error: "Page xxxxxxxx is missing".
Resolution: Code changed to save the page with correct PageID and MenuID.
Reproduction:
1) Open a CA eHealth Router Page.
2) Edit a tab.
3) Save the tab.
4) Reopen the context page.
5) An error message will display.
(2.3, 182807, 21283377-01)

Symptom: Data Aggregator may stop processing all poll data.
Resolution: Data Aggregator will continue to process poll data upon the arrival of late arriving poll data.
Reproduction: Late arriving poll data due to a late SNMP poll response will cause Data Aggregator to stop processing polled data.
(2.3, 183528, 21316608-1)

Symptom: CA Performance Center performance slows and events eventually stop being forwarded to Spectrum.
Resolution: CA Performance Center will retrieve events incrementally requests events from Event Manager when CA Spectrum sends a request for events.
Reproduction: Integrate CA Spectrum with a substantial number of devices as a data source and export multiple landscapes.
(2.3, 184355, 21289606)

Symptom: Data Collectors consume 100% CPU and disconnect from Data Aggregator.
Resolution: Improvements to Data Collector where made to correct an ICMP congestion problem.
Reproduction: Perform a large inventory discovery with ICMP Discovery enabled.
(2.3, 185176, 21323393-1)

Symptom: Unable to navigate to the original/real group from a reference group in the Group admin UI of CA Performance Center.
Resolution: The link used to navigate to the original/real group in the Group admin UI of CA Performance Center now works.
Reproduction:
1) Navigate to the Group admin UI in CA Performance Center.
2) Create a new user or site group.
3) Make a copy/reference to that newly created group.
4) Navigate to the copy/reference of the newly created group.
5) When clicking the link that is supposed to take you to the original/real group that is being referenced, nothing happens.
(2.3, 187947, 21331299-1)

Symptom: When configuring email, user has to select both SSL and Authentication options.
Resolution: Separated SSL and Authentication options. User can select either option when configuring email server.
Reproduction:
1) Configure email.
2) User cannot just select SSL option and not Authentication option.
(2.3, 191523, 21349710-1)

Symptom: In CA Performance Center, the browser's back button could go too far back and remove a selected group.
Resolution: The URL now encodes the group id when appropriate, via the GroupID parameter.
Reproduction: Select a Dashboard and change the group. Then select an item on a view in the dashboard to drill into its page. Now, click the browser's back button, and you will see that it goes back out to the Dashboard, but the group is not the one that you had changed to.
(2.3, 194692, 21295191-1)

Symptom: DeviceTypes.xml file overwritten at upgrade of Data Aggregator.
Resolution: Updated documentation for upgrade details on how to backup and restore file.
Reproduction: Apply changes to DeviceTypes.xml file, upgrade Data Aggregator; changes no longer appear in DeviceTypes.xml file.
(2.3, 194698)
(2.3, 204397)
(2.3, 194879)

Symptom: Certain files on Data Aggregator and Data Collector can be overwritten during upgrade.
Resolution: Documented process for back and restoration of customized files.
Reproduction: Upgrade Data Aggregator or Data Collector after customizing JMS or Discovery Configurations.
(2.3, 194699)
(2.3, 195436)

Symptom: Reporting times may degrade when an Event rule is enabled for a monitoring profile.
Resolution: Orphan event cleanup process was improved to be more efficient and run less frequently.
Reproduction: Create an Event rule in a monitoring profile that is applied to a collection.
(2.3, 194877, 21361085-1)

Symptom: VMWare ESX Host tab/view does not have the Edit Tab or Delete Tab option.
Resolution: Fixed the VMWare ESX Host tab to enable edit and delete of this tab.
Reproduction: Navigate to VMWare ESX Host tab/view, right-click the tab, the right-click menu doesn't have the Edit Tab or Delete Tab option.
(2.3, 195595)

Symptom: Monitored Devices search in Data Aggregator admin view does not return.
Resolution: Code changes to improve large scale searching performance, to make sure results are displayed in a reasonable amount of time.
Reproduction: On large scale systems, searches may periodically not return results.
(2.3, 195917, 21356530-1)

Symptom: View drill-down from CA Performance Center to CA Network Flow Analysis uses incorrectly formatted start and end times.
Resolution: Replace the link template with time format recognized by CA Network Flow Analysis.
Reproduction: Drill down from a CA Performance Center view to the CA Network Flow Analysis source (such as from the IP Performance tab of an interface, stacked protocol trend view).
The hyperlink contains the start and end times in the proper format that CA Network Flow Analysis will not recognize and CA Network Flow Analysis will use 24-hour time period.
(2.3, 196133)

Symptom: After changing the default SNMP retries and timeout via Data Aggregator rest webservice, the change takes effect. However, after Data Aggregator reboots, the change is lost.
Resolution: Removed the part of code which reset SNMP Profile during Data Aggregator starts.
Reproduction: Update SNMP Profile's retries/timeout via <a href="http://yanya04-rh55vm1:8581/rest/profiles/">http://DA:8581/rest/profiles/</a>. Restart Data Aggregator. Check the SNMP Profile via the same link. Then the change is lost.
(2.3, 196681, 21367622-1)

Symptom: Data Aggregator synchronization with CA Performance Center fails.
Resolution: This issue is resolved when daily delete_vector purging script is configured and deployed for Data Repository.
Reproduction: Occurs due to database access contention due to large number of record deletions.
(2.3, 196928, 21367624-01)

Symptom: On the "Unified Communications - Worst performance" dashboard, if you go to the help page by clicking any "?" mark and go to the search tab, you get a 404 error if you then click the "Click here to search the bookshelf" link.
Resolution: Removed the "Click here to search the bookshelf" link.
Reproduction:
1. Open the CA Unified Connections Monitor help by clicking '?" in page "Unified Communications - Worst performance".
2. Switch to Search tab.
3. Click the link "Click here to search the bookshelf".
(2.3, 197698)

Symptom: IP subnet mask incorrectly defined for an interface introduces synchronization failure.
Resolution: Code change to prevent invalid mask from impacting data source synchronization.
Reproduction: Not applicable - occurs with misconfigured environments.
(2.3, 199544, 21379504-01)

Symptom: The line colors for the Latency Trend report are too light to be seen easily.
Resolution: Modified the light blue trend line to be a bit darker so it can more easily be seen.
Reproduction:
1) Go to the page builder.
2) Click the "Select Context" link.
3) Set the "Context Type" to Device and select a Router as the context item.
4) Click Ok.
5) Save the page.
The lines that appear on the view may be more difficult to read than other trend based views.
(2.3, 200343, 21354403)

Symptom: High and/or increasing CPU utilization on the Vertica Data Repository (v6.0.1) system, which manifests itself in performance problems in other CA Infrastructure Management components, such as Data Aggregator, and Data Collector. For example, this issue has been reported as a synchronization failure between Data Aggregator and CA Performance Center, increased Dashboard loading times in Performance Center user interface, and so forth.
Resolution: Upgrade Vertica to 6.0.2-0 which contains a fix for the Analyze Statistic performance problem.
Reproduction: Occurs over time as Data Aggregator is running and collecting data.
(2.3, 201093, 21387025-1)
(2.3, 200396)

Symptom: Data Collector is restarted but not connected to Data Aggregator.
Resolution: Rearrange the startup sequence for Data Collector's bundles, making sure the library file starts before dependent bundles start.
Reproduction: It is a timing issue that happens randomly when Data Collector starts.
(2.3, 202345, 21391066)

Symptom: Data Collector that runs Element Management System for Starent Device Pack is restarting on its own.
Resolution: Disable Element Management System Device packs' functionality to call "System.exit()".
Reproduction: Run Starent Device pack, let it process an empty file.
(2.3, 204028, 21395928-1)

Symptom: After editing a context page while administering a tenant, navigating to the page gives a "context menu XXXXXXX is not available" error, where XXXXXXX is a numeric internal page code.
Resolution: Modified code so when you administer a tenant, but do not proxy a user, the option to reorder tabs is not available. It will be available when you proxy a user of that tenant who has rights to edit context pages. Also modified the page code to choose your tab order rather than the tab order of the tenant when you are administering.
Reproduction: Log in to the system with a user who has rights to administer tenants and edit context pages. In the menu, Go to "Admin" -> "Tenants". Select a Tenant with discovered items and click "Administer" (Do not proxy a user afterwards). In the menu, go to "Inventory" -> "Devices". Select any device with a context page (for example, Router, Server, or Switch). On the current tab, select the "gear" icon and choose "Reorder Tabs". When the dialog appears, change the order of the tabs and click OK. The page will refresh with the above "context menu XXXXXXXX is not available".
(2.3, 206065, 21392715-1)

Symptom: Vertica 6.0.2 normal operation of a cluster may result in CPU spikes on low-end devices that are in the same subnet as the nodes in the cluster, and/or may be detected by broadcast storm detection filters.
Resolution: Follow these instructions to reconfigure Vertica to use point-to-point communication between the nodes in cluster: Manually re-configure Vertica to use point-to-point communication between nodes instead of broadcast communication. This can be done by:
1. Bring down your Data Aggregator and Data Repository - On Data Aggregator: /etc/init.d/dadaemon stop - Stop the databases on all nodes in the cluster: As Data Repository Administrator, launch /opt/vertica/bin/admintools, select (4) Stop Database - Wait for all the database nodes to be stopped. You can confirm by selecting (1) View Database Cluster State in admintools
2. On any one node in the cluster, you must run the install_vertica command identical to the one when the cluster was originally set up but with one addition - the "-T -S default" option should be included to basically specify that spread should use a point-to-point configuration like this:/opt/vertica/sbin/install_vertica -s host1,host2,host3 -r /opt/CA/IMDataRepository_vertica6/vertica-6.0.2-0.x86_64.RHEL5.rpm -u <DR_ADMIN_USER>-l <DRADMINS_HOME_DIRECTORY> -L <VERTICA_LICENSE> -T -S default So for example., if your original install_vertica command was run as:/opt/vertica/sbin/install_vertica -s host1,host2,host3 -r /opt/CA/IMDataRepository_vertica6/vertica-6.0.2-0.x86_64.RHEL5.rpm -u dradmin -l dradmin -L /opt/CA/IMDataRepository_vertica6/vlicense.datthen the new command would be:/opt/vertica/sbin/install_vertica -s host1,host2,host3 -r /opt/CA/IMDataRepository_vertica6/vertica-6.0.2-0.x86_64.RHEL5.rpm -u dradmin -l dradmin -L /opt/CA/IMDataRepository_vertica6/vlicense.dat -T -S default
Reproduction: Occurs under normal operation of the CA Infrastructure Management Data Repository.
(2.3, 206094, 21400575-1 )

Symptom: Unable to discover Data Aggregator device that is running an SNMP agent.
Resolution: Code changes allow for both SNMP monitoring and self-monitoring of Data Aggregator to coexist.
Reproduction: Create a discovery profile with the name of IP address of Data Aggregator. Note that it cannot be discovered.
(2.3, 206163, 21408631-1)
(2.3, 206236)

Symptom: Events from CA Infrastructure Management stop appearing as alarms in CA Spectrum.
Resolution: Code changes in CA Spectrum resolve this issue. PTF_9.2.37 and CA Spectrum r9.3 include the resolution.
Reproduction: Not applicable - no simple scenario causes this, it happens in larger environments.
(2.3, 206279, 21395668-1)

Symptom: The left top box of the Discovery history GUI does not show the discovery instance date info. Only the radio button is present.
Resolution: Fixed the Japanese Language resource file.
Reproduction: The problem reproduced only on the CA Performance Center installed in Japanese Language. English language works fine.
(2.3, 206688, 21376575-01)

Symptom: Issue with CPU/Memory Display Dashboard whereby all ASR routers show an Average, Maximum and Minimum utilization as 99.98%.
Resolution: We now additionally filter out mem components whose cempMemPoolType is "other" and cempMemPoolName contains the string "lsmpi_io".
Reproduction: Poll EnhancedMemPool components on cisco ASR routers and view memory utilization dashboard.
(2.3, 207482, 21236156)

Symptom: Data Aggregator or Data Collector processes encounter an exception after running for a period of time - depending on the level of scale. This exception text can be seen in the corresponding karaf.log: "The use count for the service overflowed".
Resolution: The problem has been resolved by patching the OSGI container's code that was throwing the exception. Data Aggregator and Data Collector processes should continue to run without encountering this exception in any of their internal processing.
Reproduction: Problem is seen in due course of time. In a large scale environment, it was seen once in 7 days on Data Aggregator, and once in approximately 15 days on Data Collector.
(2.3, 207509, 21404933-1)

Symptom: When doing a silent install of Data Collector, the install completes successfully, but Data Collector does not work. This is because the host name is missing in the local-jms-broker.xml.
Resolution: Fixed the silent install for the Data Collector installer. It now installs the local-jms-broker.xml file with the correct host name.
Reproduction: Follow User Guide instructions for installing the first Data Collector and having it create an installer.properties file. That command, is: "./install.bin -r"
That creates an installer.properties file and after that first Data Collector install, the host name is correct in the local-jsm-broker.xml file. That file is found in /opt/IMDataCollector/apache-karaf-2.1.3/deploy/. The section of concern with the host name is:
<networkConnectors> 
 <networkConnector name="manager"
 uri="static:(tcp://polmi05-VM71157:61616)" duplex="true"
 suppressDuplicateTopicSubscriptions="false" />
</networkConnectors>

The instructions in the guides then direct you to install the secondary and other Data Collectors using the installer.properties file with this command:
" ./install.bin -silent -f <path_to_installer.properties_file>"

After running that using the installer.properties file created with the first command per the guide, the same section in the local-jsm-broker.xml file looks like this:
<networkConnectors>

<networkConnector name="manager"
 uri="static:(tcp://localhost:61616)" duplex="true"
 suppressDuplicateTopicSubscriptions="false" />
</networkConnectors>

It should have the host name of Data Aggregator per the installer.properties files content in the uri field, not localhost.
(2.3, 208031, 21405926-1)

Symptom: Data Aggregator data source synchronization failure seen in CA Performance Center Data Sources page. Data Aggregator or Data Collector processes encounter an exception after running for a period of time - depending on the level of scale. This exception text can be seen in the corresponding karaf.log: "The use count for the service overflowed".
Resolution: The problem has been resolved by patching the OSGI container's code that was throwing the exception. Data Aggregator and Data Collector processes should continue to run without encountering this exception in any of their internal processing.
Reproduction: Problem is seen in due course of time. In a large scale environment, it was seen once in 7 days on Data Aggregator, and once in approximately 15 days on Data Collector.
(2.3, 208071, 21419282-1)

Symptom: This customer uses this Infrastructure Management environment to monitor a single secure network full of about 150 Cisco 6500 devices. The devices have a variety of firmware/IOS installs. The problem is seen for a random set of about half the devices; approximately half the devices in the environment show no Availability data. When accessing the devices Router Pages Details tab view, the Availability Calendar Trend and Availability Heat charts both show 'No Data Available' messages for problem devices.
Resolution:
1. Enhanced the Data Collector debug page to list any devices that have a mismatch of DcmID between the device item and its associated polling configuration. This will help identify devices associated to the stale Data Collector.
2. While looking at polling configuration by IP address, the Data Collector debug page will now also show the DcmID of the polling configuration item.
Reproduction: Unclear, but there was a misstep during Data Collector upgrades previously, and there was also an invalid migration of Data Collector from one host to another (without following the instructions outlined in the document).
(2.3, 210786, 21424568-1)
(2.3, 210785, 21430900-1)

Symptom: Data Collector is not assigned to Default Domain after silent install.
Resolution: Fixed the silent install for Data Collector installer.
Reproduction: Run console install to generate an installer.properties file. Run silent install with this file.
(2.3, 211221)

Symptom: When you use self-certification wizard to create a Vendor certification, ipaddress type in the mib is mapped to OctetString.
Resolution: Fixed the code to map the mib's ipaddress type to correct type.
Reproduction: Import a mib which has ipaddress type, and use the certification wizard to create a vendor certification.
(2.3, 211440, 21310554-01)

Symptom: The 3x2x1 layout option is missing for custom dashboards.
Resolution: The 3x2x1 layout option will now be present when editing or creating a custom dashboard.
Reproduction: Go to a Routers "Router Pages" dashboard view. Select the System Health tab for it. Here we see charts laid out as follows: 3 in first row, then 2 in second row, 2 in third row, and 1 in fourth row. Try to edit that tab to create a custom one and change the data in its charts, and you are presented with a list of layout options that is missing the 3x2x1 option.
(2.3, 214190, 21443417-1)

Symptom: Performance View launch to CA Performance Center from CA Spectrum alarm fails.
Resolution: Code changes help ensure proper URL is provided as part of event notification which produces CA Spectrum alarm containing the URL.
Reproduction: Can occur with Spectrum alarms where CA Performance Center has multiple network interfaces.
(2.3, 214228, 21345726-1)

Symptom:
1 - Unable to associate a Data Collector to a Tenant or Domain more than once.
2 - Change of Tenant name associated to Data Collector causes Data Collector to fail to run.
Resolution: Code changes help ensure changes to tenant and IP Domain name are updated for each Data Collector that is affected.
Reproduction:
- Create a new Tenant named 'Remote Tennant'.
- Assigned it to a Data Collector.
- Realize the mistake in the spelling of 'Tennant' and fix the name of the Tenant in CA Performance Center.
- In the Data Aggregator Data Collector list, note the name is listed correctly against Data Collector after the correction, showing 'Remote Tenant'.
- Restart Data Collector and it will fail to run until the name is also manually corrected in the com.ca.im.dm.core.collector.cfg file.
(2.3, 215001, 21432976-1)

Symptom: When looking at "Interfaces Over Threshold" view, the label for the "Percent Time Critical" column will show greater 75%, when it should be showing 75%.
Resolution: Label construction and formatting was fixed to not touch the value that is entered and the percentage values now display correctly.
Reproduction: Install CA Performance Center 2.2 and add CA Network Flow Analysis 9.1 SP2 as a data source. Navigate to "Interfaces Over Threshold" view and look at column header labels. Two show the wrong headings, stating > 0.75% or > 0.50%.
(2.3, 215312, 21403021-01)

Symptom: For some stacked trend charts in non-Data Aggregator data sources, the top of the chart could be chopped off and not show any values greater than 70%.
Resolution: Modified the code to calculate the highest percentage value over the time period for the report and use this value as the maximum that should be displayed. This allows for values greater than 70% to display properly.
Reproduction: You must have data such that the sum of all the values in a time interval is greater than 70%. In this case, the chart will only show values up to 70% even though your total percent could be higher than this.
(2.3, 215366, 21451833-1)

Symptom: There are duplicate interface components in CA Performance Center's monitored device view, one is 'active' and polling, the other is 'active' but 'not polled'. This happened because the interface was somehow unavailable when Automatic Change Detection was running causing CA Performance Manager to retire the interface. This interface was then available again the next time Automatic Change Detection ran.
Resolution: The UI will now display retired interfaces correctly. Going forward, CA Performance Manager will re-activate the retired interface when it runs into this problem instead of creating a new one.
Reproduction: Not applicable.
(2.3, 217413, 21467821-01)

Symptom: The charts in the center column of the page builder 1x3x1 layout are taller than left/right columns.
Resolution: Changed the layout of the three columns in the 1x3x1 layout to all be 33% (i.e. changed center column from 34% to 33%). Only whole numbers are supported and this means the columns add up to 99% of the width of the page, leaving a slight space at the right, but this is consistent with what the other 3-column layouts do, and looks much better than the too-tall center column did.
Reproduction: Create a new dash board using the 1x3x1 layout. Add any three views to the center section, the section of the page that displays views three across. The dashboard may show different heights for the rectangles.
(2.3, 218116, 21337365-01)

Symptom: In the Google Chrome web browser, when viewing the "Tree View" tab for Monitored Inventory > monitored devices in the Data Aggregator web administration interface, there is a hidden scroll bar to see all devices. The issue is only seen when using Google Chrome. Internet Explorer and Firefox do not replicate the issue.
Resolution: Modified the css for this page element to hide overflow, which causes Google Chrome to not display the extra scroll bars.
Reproduction:
In the Google Chrome web browser, navigate to the Data Aggregator user interface pages. Navigate to Monitored Inventory and select the Monitored Devices menu item. There is an extra scrollbar in the "tree view" tab that displays items by collection or monitoring profile.
(2.3, 219426, 21402368-1)

Symptom: When using custom vendor certifications which polled for data that should be a negative number, the values being returned were positive numbers (as they were being converted to unsigned integer incorrectly) causing a discrepancy between what was being displayed in CA Performance Center, and what the actual values were according to the device.
Resolution: Negative values are now supported. You can report on these values.
Reproduction:
1. Set up a simulator or device which can send a negative number as a response during SNMP polling.
2. Add custom vendor certificates which poll for that negative data.
3. View the data that is coming in and check that it has been converted to a positive number and not left a negative number, and perform an SMNP get on the device to confirm that it should be a negative number.
(2.3, 220813, 21471005-1 )