Known problems affecting InfoSphere Information Server Enterprise Hypervisor Edition V9.1

Release notes


Abstract

This document contains issues with IBM Workload Deployer and IBM PureApplication System that affect the deployment of the pattern for InfoSphere Information Server Enterprise Hypervisor Edition, Version 9.1.

Content

Problem description Problem conclusion Workaround

Extending an image to add a network interface fails with the following error:
RM14024 (Incompatible device backing specified for device'3')

This issue occurs only on standalone installations of IBM Workload Deployer. The issue is not applicable on IBM PureApplication System.
Extending an image to add a network interface is not supported. Additional network interfaces can be added during deployments by adding a NIC add-on to the pattern.

Images that are exported from IBM PureApplication System cannot be imported to a standalone installations of IBM Workload Deployer.

This problem occurs because IBM PureApplication System adds an entry for a .cloudburst file in the  OVA image name.ovf file of the OVA image. However, the .cloudburst file does not exist.

Therefore, when you use IBM Workload Deployer to import the OVA image, the .cloudburst file cannot be found and the import fails with an error.
Unless you first remove .cloudburst file entries from the .ovf file of the OVA image, you cannot import images from IBM PureApplication System to a standalone installation of IBM Workload Deployer. Before you import an exported image from IBM PureApplication System to IBM Workload Deployer, .cloudburst file entries from the .ovf file of the OVA image.

To remove .cloudburst file entries from the .ovf file of an .OVA image:
  1. Extract the exported .ova image file.
  2. Open the  OVA image name.ovf file.
  3. Remove any .cloudburst file entries from the list.. The following item is an example of a file entry for a .cloudburst file in a .ovf file:
    <ovf:File ovf:href=".cloudburst" ovf:id=".cloudburst" ovf:size="501"/>
  4. Recompress the OVA image with the modified .ovf file.
  5. Import the modified OVA image file into IBM Workload Deployer.

Importing images fails with IOError: [Errno 503] Service Temporarily Unavailable (Test Env = Rack 32).

Error:

Traceback (most recent call last):
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/patternloader.py", line 698, in JSONtoPattern
    obj = _findVirtualImage(virtuaImage, dest)
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/patternloader.py", line 166, in _findVirtualImage
    _waitForVirtualImageImport(virtualImage)
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/patternloader.py", line 147, in _waitForVirtualImageImport
    virtualImage.refresh()
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/resources/restresource.py", line 321, in refresh
    self._restattrs.refresh()
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/resources/restdict.py", line 99, in refresh
    newData = self._munge(http.get(self._uri))
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/http.py", line 331, in get
    return _httpRequest(uri, 'GET', headers = _defaultHeaders(), responseHandler = responseHandler, externalServer = externalServer)
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/http.py", line 257, in _httpRequest
    return responseHandler(resp)
  File "/mnt/disk/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/http.py", line 146, in _defaultResponseHandler
    raise IOError(resp.status, utos(resp.reason))
IOError: [Errno 503] Service Temporarily Unavailable


Image import failed with error: "CWZIP1906E Failure during the extraction of image binary".

Error:

Time elapsed: 12 minutes ...
Time elapsed: 12 minutes ...
Traceback (most recent call last):
  File "/dsdata/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/patternloader.py", line 698, in JSONtoPattern
    obj = _findVirtualImage(virtuaImage, dest)
  File "/dsdata/AIOCN/pure.cli/lib/1.0.0.4-201302252130011/deployer/patternloader.py", line 188, in _findVirtualImage
    assert len(matches) == 1 and matches[0].name == name, utils.utos(message('IWD00023') % (name))
AssertionError: No virtual image named, "IBM InfoSphere Information Server Enterprise Node Hypervisor Edition 9.1 for RHEL Server" is defined.
As per the documentation in the following link: Importing the InfoSphere Information Server virtual system pattern, if you follow CLI method to import Information Server using PureApplication System (Intel 1.0.0.4 and IWD 3.1.0.7), it fails with the errors. It blocks importing images but not the pattern and script packages. Follow these steps:
  1. Add Information Server Virtual Images to IBM Workload Deployer using the Workload Console as described in Adding virtual images to the catalog.
  2. After the images are imported, you can proceed with CLI import procedure as described in, Importing the InfoSphere Information Server virtual system pattern, to import the pattern and script packages together.

Cloned virtual machine is not responding after the hot update operation on the stand alone IBM Workload Deployer.

Error:

[2013-03-23 18:30:59:915 UTC] 00006786 RXA E com.ibm.tivoli.remoteaccess.SSHProtocol beginSessionImpl [9.37.134.212] CTGRI0001E The application could not establish a connection to 9.37.134.212 .
[2013-03-23 18:31:26:584 UTC] 00006787 RXA E com.ibm.tivoli.remoteaccess.SSHProtocol beginSessionImpl [9.37.134.212] CTGRI0001E The application could not establish a connection to 9.37.134.212 .
[2013-03-23 18:31:26:584 UTC] 00005530 RXA E com.ibm.tivoli.remoteaccess.BaseProtocol beginSession [9.37.134.212] CTGRI0026E A connection could not be completed to 9.37.134.212 during the specified timeout interval.
[2013-03-23 18:31:59:183 UTC] 00006790 RXA E com.ibm.tivoli.remoteaccess.SSHProtocol beginSessionImpl [9.37.134.212] CTGRI0001E The application could not establish a connection to 9.37.134.212.
The cloned virtual machine does not respond if you have followed these steps:
  1. Deploy the InfoSphere Information Server Enterprise Hypervisor Edition V9.1 for Red Hat Enterprise Linux Server Engine Cluster Configuration pattern.
  2. Clone the Compute node virtual machine to expand the Compute node instance.
  3. Stop the cloned instance to increase the CPU and memory.
  4. Modify or increase the CPU and Memory.
  5. Start the Cloned virtual machine.


While deploying this pattern, select the required number of CPU and Memory to avoid the issue.

PureApp 1.0.0.4 License management is failing in following cases:
  • insufficient licenses - ignore
  • insufficient licenses - enforce
  • insufficient licenses - warn
  • virtual system History logs
Unable to track PVU usage correctly on deployed instance.
  • Upgrade to IPAS v1.1(IWD 4.0).
  • Check APAR IC92961 for fix and apply.

PureAS 1.0.0.4 Importing Service Level Upgrade image failed with errors.

Error:

f0a84e8e-2326-42b3-82aa-3ab9e9bd4a6c [06-03-13 16:04:26] 0043 com.ibm.pdq.runtime.exception.DataRuntimeException: [pdq][10107][2.7.116] An error prevented the update operation from completing successfully.;  Caused by: java.sql.SQLException: The specified table (ed295535fa4e846e1b3dc34baa68fe0d0) is not in the database.
Unable to apply IS 9.1 fixpack through IWD Service Upgrade Level operation.
  • Upgrade to IPASv1.1.
  • Use e-fix if the IS9.1 iFix is less than 2 GB.
  • Install IS9.1 iFix manually on the deployed instances.

Deployment failed due to 'Default AIX add disk' execution failure.

Error:

The required initialization script package Default AIX add disk failed to execute successfully on virtual machine.

Error:

Unable to extend the file system to max size.
PureAS 1.0.0.4 Deployments will fail if using the 'Default AIX add disk' option to add disk to an existing volume group to VM. With PureAS 1.0.0.4, one can use 'Default AIX add disk' to add extra storage volume to VM with a new volume group. If use the 'Default AIX add disk' option to add disk to an existing volume group, please upgrade to IPASv1.1 as the expand storage volume function is a PureAS 1.1 feature, not available for PureAS 1.0.0.4.

Add PID button is enabled and did not allow to add PIDs when virtual system image is deployed.

Error:

The resource could not be updated.
Unable to add PID when virtual system image is deployed.
  • Ignore Add PID button, when virtual system image is deployed.
  • Upgrade to PureAS v1.1, where the Add PID button is disabled.
  • Delete all the virtual system instances and virtual system patterns under virtual system image to add PID to that image.
  • Issue 1: Virtual System Pattern Deployment hangs at stage "Starting virtual machine...'
  • Issue 2: Last stage of above issue leads to deployment failures with the following error.

Error:

Some virtual machines within the virtual system did not initialize correctly.
AE.sh-related processes (including the one starting maestro) were waiting for the file /opt/ibm/ae/lock/hw-lock/pid to be cleared, however, it was not cleared possibly because the process that was supposed to unlock or clear it was somehow terminated (this thought popped up since the pid file had no content (size=0) while it was supposed to contain a process ID). This caused an infinite wait.
  • Issue 1: Login to VM of the Virtual System Instance which takes more than actual deployment time. If the size of file /opt/ibm/ae/lock/hw-lock/pid is zero, you need to delete it. Then, your deployment process will continue.
  • Issue 2: If you have not noticed deployment is taking too long time, then your deployment may fail after an infinite wait. Stopping and then starting Virtual System instance will make your instance available as "Virtual system is ready". But attached Script Packages will not be executed. You may have to deploy new instance to get it executed.
Use latest Power Patterns (v1.0.0.3) to avoid deployment hanging issues.

In a Engine Cluster Configuration Pattern deployed instance, if a Compute Node Virtual machine is deleted, when it is in stopped state, the "IBM InfoSphere Information Server 9.1 Compute Node Delete" script package will not get executed. This might lead to deleted VM host entries in /etc/exports and default.apt. This may cause unexpected results during DataStage Job Executions.
Script Packages cannot be executed if virtual machine is in stopped state and then deleted. Virtual Machines need to be started before deleting.

  1. Before deleting a Compute Node Virtual Machine, ensure that the Compute Node Virtual Machine is live and started.
  2. If you have deleted the Compute Node in stopped state, you can manually add the short-host name of the deleted node to /opt/IBM/IIS-Registry/DeleteNodes (ie... touch /opt/IBM/IIS-Registry/DeleteNodes/<shortHostName>) and re-run the Conductor Node Update Script Package.

Related information

System requirements for Hypervisor Edition V9.1
Installing Hypervisor Edition
Architecture for IBM private cloud
Quick Start Guide
Download Hypervisor Edition V9.1

Rate this page:

(0 users)Average rating

Add comments

Document information


More support for:

InfoSphere Information Server

Software version:

9.1

Operating system(s):

AIX, Linux

Reference #:

7036943

Modified date:

2014-01-19

Translate my page

Machine Translation

Content navigation