Deployment Patterns and Anti-patterns when using Cloud Services

Authors: PaulEllis
Build basis: Engineering Lifecycle Management 6.0.x

This article documents issues that we see when deploying the Engineering Lifecycle Management software on Cloud Services such as: Microsoft Azure; Amazon Web Services; IBM Cloud; Google; and others. This article will not cover the issues from ClearObject.

The Server Virtualization Policy for IBM Software states:

"In response to the significant increase in Cloud and PureSystem technologies now available to its clients, IBM has established a set of virtualization environments for servers that we will support across our multi-platform product portfolio for servers. These virtualization environments consist of the operating system and hypervisor technology. This information is intended to help clients plan their strategic deployment environments for applications and achieve the economic benefits available through server consolidation."

As you can see from the jazz.net forum post from our Offering Manager Daniel Moul, we have stated:

Yes. CLM is supported when running in Hyper-V, which is used in the Azure environment.
Note that if using Azure you will still need to bring perpetual licenses and provide admin services.
If you would rather move to SaaS and/or have someone take care of the administrative services, you could move to and IBM cloud offering. To learn more see the following:
https://www.ibm.com/marketplace/cloud/application-lifecycle-management
https://www.ibm.com/marketplace/cloud/engineering-solutions-on-cloud/"

This statement does not state that we support every aspect that a Cloud provider can offer. There is still a very strong expectation that the system will meet our existing IBM Engineering Support T&C, and specifically ALL OSs and middleware used within the Cloud Service meets our Systems Requirements, for the specific version of IBM Engineering that is deployed.

Windows restarts

If Windows patches are set to deploy automatically, you MUST shut down the indices and CLM services, cleanly, before this is scheduled in case of the Windows patch requiring a restart.

Amazon Web Services Relational Database Service - Oracle

When creating Oracle tables using the commands that are prescribed in the Knowledge Center or Interactive Installation Guide, on Amazon RDS, you may encounter the following issue(s).

"The data warehouse tables could not be created. Try the action again.ID CRJAZ1745E

CRRTC8030E The creation of the data warehouse failed.

com.ibm.team.repository.common.TeamRepositoryException: CREATE TABLESPACE VNF_IDX LOGGING DATAFILE '/rdsdbbin/oracle/VNF_IDX_1.ora' SIZE 200M AUTOEXTEND ON NEXT 50M MAXSIZE UNLIMITED EXTENT MANAGEMENT LOCAL SEGMENT SPACE MANAGEMENT AUTO"

and

java.sql.SQLException: ORA-00604: error occurred at recursive SQL level 1 ORA-20900: RDS only supports Oracle Managed Files. Check ddl and remove any named identifiers ORA-06512: at "RDSADMIN.RDSADMIN", line 160 ORA-06512: at line 2 at

* What is AWS RDS telling you?:

The version of Oracle in use within the AWS RDS is not your typical on-premise Oracle. Therefore there maybe issues you will need to work around which will not be documented within the IBM documentation.

Standard tablespace create in this way (which they must have already worked around:)
CREATE TABLESPACE JTS
DATAFILE 'ORACLE_BASE/oradata/CLMDB/JTS.DBF'
SIZE 1G AUTOEXTEND ON EXTENT MANAGEMENT LOCAL AUTOALLOCATE;

What the DW is doing:
CREATE TABLESPACE VNF_IDX LOGGING DATAFILE c:\somepath\VNF_IDX.ora'
ALTER TABLESPACE VNF_IDX ADD DATAFILE c:\somepath\VNF_IDX.ora'

Specifying the paths (emboldened) is the problem here. The DBA will need to work within the Oracle command's syntax to create these and allow the Managed File System to allocate the datafies.

Data Collection Component startup reports SQL Message: ORA-00942: table or view does not exist using Oracle server for Jazz Reporting Service

Attempts to start the Data Collection Component (DCC) are successful, but the following message "SQL Message: ORA-00942: table or view does not exist" is reported in the dcc.log after installing IBM Jazz Reporting Service on Oracle.

The DCC service reports the following in the dcc.log

11/07/2018 19:17:26,254 [Default Executor-thread-81] INFO sqlExceptionLogger  - Syntax error or access rule violation
SQL: DROP TABLE DCC_DB_USER.REPOSITORY_GTT_STRING
SQL Exception #1
SQL Message: ORA-00942: table or view does not exist
SQL State: 42000
Error Code: 942

java.sql.SQLSyntaxErrorException: ORA-00942: table or view does not exist

........
Caused by: Error : 942, Position : 23, Sql = DROP TABLE DCC_DB_USER.REPOSITORY_GTT_STRING, OriginalSql = DROP TABLE DCC_DB_USER.REPOSITORY_GTT_STRING, Error Msg = ORA-00942: table or view does not exist
    at oracle.jdbc.driver.T4CTTIoer11.processError(T4CTTIoer11.java:498)
    ... 199 more

The table name in the error above is DCC_DB_USER.REPOSITORY_GTT_STRING but the DCC is is the user name; the table name is created based on the user name and the Global Temporary Tables Type. The table name in your log may, therefore, be different.

This is a known issue discussed in the following:
478063: Silent Upgrade 606iFix006-6061RC3 Error observed in DCC logs after upgrade
478089: Internal API to drop a table should not throw SQLException if the table does not exist

These issues were initially suspected to be related to the Amazon Web Services RDS due to the alternative technique required to create the Data Warehouse. This procedure has been confirmed to still be correct and these errors are benign and can be ignored. They relate to temporary tables which are being removed and in this instance, they do not exist anyway.

Uploading models to Rhapsody Model Manager/Rhapsody Design Manager issues

When uploading a Rhapsody model with the "RTC-Client" to the Model Manager.

1. CRRTC5038W "The versioned content service temp directory has been deleted."

This error was fixed by changing the temp directory.

1. CRJAZ0098E error, after fixing 1st error

CRJAZ0098E following service not running:
com.ibm.team.scm.common.IScmService{/am1/service/com.ibm.team.scm.common.IScmService}</i>

The server returned the HTTP error 504 with the following error message:
Gateway time out.

See the Amazon Web Services help page for Timeout issue is resolved after changing timeout variables on AWS server.

Related topics: Deployment web home, Deployment web home

External links:

• IBM

Additional contributors: ValerieLampkin, MirkoHartwig, StevenBeard, GertWinderlich