Release Notes
Abstract
This release notes document lists the known issues in this release.
Content
The following issues are known in this release.
- Using Java 7 with your browser client in the PureApplication System Monitoring Portal is currently not supported in IBM PureApplication System.
-
Administering the system:
-
High availability:
-
IBM Construction and Composition Tool:
-
Managing virtual systems and virtual applications:
- WebSphere Application Server hypervisor image hangs on capture
- Automatic registration of database monitoring is not working in virtual systems
- Increasing the number of CPUs for deployed patterns causes WebSphere Application Server to stop responding
- The Revert function does not work after the virtual application upgrade
- Incorrect values for the virtual CPU in-use limits in the environment profile
- The virtual memory utilization of some virtual machines (LPARs) show 0% in user interface
Administering the system:
The compute node is quiesced because RMC is inactive
| Problem: | Error: CWZIP6190E The compute node: ${0} has been quiesced because one (or both) of the VIOS do not have active RMC connections ${1} While the compute node is quiesced workloads continue to function, but further deployments are prevented. |
| Resolution: | Wait for the compute node to automatically become available again. If after 30 minutes the compute node is still quiesced due to an inactive Resource Monitoring and Control (RMC), perform the following steps
|
Critical events are occurring during rack shutdown or startup
| Problem: | Critical events are generated by PureSystems Manager during rack shutdown and/or startup |
| Resolution: | During rack shutdown, you can ignore critical events, such as:
During rack startup you can ignore the following error type:
|
Parallel deployments to a new compute node will boot with broken networking
| Problem: | Running two or more (n) parallel deployments to a new compute node will likely result in n-1 virtual machine instances booting without valid networking. The first virtual machine deployment to a new compute node is the trigger that configures the hypervisor's networking (VLANs, etc...). While the first deployment configures hypervisor networking, the other parallel deployments fail to block until this step is completed. As a result, subsequent virtual machine instances run with invalid network configurations. |
| Resolution: | After adding a new compute node to a cloud group, perform a single virtual machine deployment to the compute node. Once the first compute node deployment is complete, the hypervisors for the compute node are properly configured and parallel deployments can now be completed. Note: These instructions will apply also, if the compute node is removed from the cloud group and added to a new cloud group. |
A connection to the database server failed
| Problem: | Cannot create new cloud groups when the PureSystems Manager is powered off or unavailable |
| Resolution: | Determine if the database is down by running the following command: <sudo dmidecode | grep "System Inf" -A 4 | grep "Serial Number" | awk '{print $3}' | grep -c $(sudo head /data/config/informix/informix.config -n 1 | cut -d"\"" -f2 | cut -d"x" -f2)> If the database is determined to be down, restart the database. |
Slow SSH connection to AIX LPARs in isolated DNS Environments
| Problem: | Due to the way OpenSSH interacts with AIX, an invalid DNS query will be generated by the LPAR's SSH daemon when a client attempts to connect to the LPAR. The DNS server will attempt to resolve the invalid query by contacting a root server. When DNS is configured in an isolated environment and no root server can be reached, either directly or by means of forwarding the request, the DNS query will be forced to time out rather than fail immediately. The SSH client will not make a connection to the LPAR until the DNS timeout has occurred. |
| Resolution: | Solution 1) Enable forwarding on the DNS server to a DNS server that has access to root servers by. adding the following configuration in the Options section In BIND's named.conf file: Note: "x.x.x.x" is the IP address of the DNS server that requests will be forwarded to: options{ ... forwarders { x.x.x.x; }; ... } Restart the BIND service after adding the configuration above. Solution 2) Disable recursion in order to reject unknown queries, instead of trying to contact root DNS servers by adding the following configuration in the Options section in BIND's named.conf file: options{ ... allow-recursion { none; }; ... } Restart the BIND service after adding the configuration above. |
High availability:
The primary PureSystems Manager always pushes workload data to overwrite the secondary PureSystems Manager
| Problem: | If the primary PureSystems Manager goes down, the secondary PureSystems Manager automatically takes over as the primary PureSystems Manager. When workload changes occur (such as deploying new workloads) only the secondary PureSystems Manager knows about the new workload changes. If the secondary PureSystems Manager goes down, and if the original PureSystems Manager then recovers and resumes its role as the primary, it might overwrite the secondary PureSystems Manager data with older information. As a result, any workload changes that were made during the outage might be lost. |
| Resolution: | If both PureSystems Managers are down, start the PureSystems Manager that was most recently functional first. |
IBM Construction and Composition Tool:
Bundles are not added to new images
| Problem: | When creating an image from a template, bundles that do not have an installation step defined are not added to the new image. |
| Resolution: | After you create a new image from a template, use the regular procedure to add additional bundles to the image. |
The PowerVM Express cloud provider is not supported
| Problem: | IBM Image Construction and Composition Tool does not support the creation of storage copy services (SCS) raw images by using the PowerVM Express cloud provider. |
| Resolution: | Use the PureApplication System cloud provider to create the image. |
Problems occurs when deploying IBM WebSphere Application Server Hypervisor Edition for PowerVM images with Image Construction and Composition Tool
| Problem: | You might notice that your deployment process stops responding with incomplete configurations in the following cases:
This problem is a current limitation for IBM WebSphere Application Server PowerVM images only. |
| Resolution | You can use the PureApplication System console to successfully clone and extend images. |
Managing virtual systems and virtual applications:
WebSphere Application Server hypervisor image hangs on capture
| Problem: | While capturing a WebSphere Application Server virtual image, you might encounter a process hang if you wait more than 60 minutes to capture the image after the initial extend has completed. |
| Resolution: | Perform the image capture within 60 minutes after you extend the image. |
Automatic registration of database monitoring is not working in virtual systems
| Problem: | When a virtual system with a DB2 database is deployed, the information in the ovf-env.xml file, used for automatically registering the database with the monitoring agent, is not properly discovered. Since the registering agent does not find the file, the database is not automatically registered. |
| Resolution: | The database needs to be manually registered using the database performance monitor. |
Increasing the number of CPUs for deployed patterns causes WebSphere Application Server service to stop responding
| Problem: | If you deploy a virtual application pattern or virtual system pattern that is configured with one CPU, and later increase the number of CPUs, WebSphere Application Server stops responding to requests. |
| Resolution: | To work around the problem, choose one of the following options:
|
The Revert function does not work after the virtual application upgrade
| Problem: | After upgrading the virtual application, the Revert function does not work from the virtual application detail web page. |
| Resolution: | Connect to the virtual machine manually and use the IBM Installation Manager silent mode function to complete the rollback of the virtual application. |
Incorrect values for the virtual CPU in-use limits in the environment profile
| Problem: | If a virtual machine is reconfigured to increase the size of the virtual CPU count while the virtual machine is in the Started state, the CPU in-use limit in the environment profile is not incremented to reflect the new size of the virtual CPU. This causes the CPU in-use limits to be incorrect. In some cases the CPU in-use limit might be negative as the value is decremented by the new size of the virtual CPU count when the virtual machine is stopped or destroyed. As a result of the incorrect CPU in-use limits, deployments exceeding the configured limits of the environment profile are not enforced correctly. |
| Resolution: | Discontinue use of the existing environment profile with the incorrect values, create a new environment profile and redeploy the instances to the new environment profile. |
The virtual memory utilization of some virtual machines (LPARs) show 0% in user interface
| Problem: | AIX operating system blocks UDP packets on port 161 that AIX SNMP agent watches on |
| Resolution: | 1. Download the emergency fix from Fix Central 2. Create a new emergency fix using the downloaded file (Catalog > Emergency fixes) 3. Select the images that need to apply the emergency fix 4. Find the problematic virtual system instance that the virtual machine belongs to 5. Find the pattern the virtual system is deployed from 6. Open the pattern and find the image name 7. Apply the fix to the image, from the Emergency fix page 8. Open the virtual system instance page, apply the emergency fix by clicking the Service link |
Related:
IBM PureApplication System release notes by version
[{"Product":{"code":"SSM8NY","label":"PureApplication System"},"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Component":"--","Platform":[{"code":"PF002","label":"AIX"}],"Version":"1.0","Edition":"","Line of Business":{"code":"","label":""}}]
Was this topic helpful?
Document Information
Modified date:
17 June 2018
UID
swg27037828