Presented by Phillip Stanton, Advisory Software Engineer for ITSM, Netcool Technical Support
Can we make changes to the WebGUI - while clustered - that does not require disjoin nodes - uninstall cluster - modify each node - reinstall cluster - rejoin nodes?
- Yes, you can change these things: users, roles, tools, filters, views, pages, navigation, theme, widgets, maps, and all the other DASH/WebGUI items.
Is there a way to export and import from the cluster - while clustered (without disjoin-uninstall-import-join)?
- I'm not exactly what is being asked here, but I'll assume that you have a production server that is in a cluster and a dev/staging server that is not and you are looking to refresh the dev/staging servers. If that is the case, no the tipcli export commands don't work while the server is clustered.
Is there a way to directly edit the database which contains the configuration for the cluster?
- Not in a way that support would support, but if you are a db2 guru I won't stop you.
Is there a way to make changes to users, roles, tools, filters, views, pages, navigation, theme, widgets, and maps that will propagate across the cluster.
- Yes, this is the purpose of a cluster.
Can we apply fixpacks without disjoin-uninstall-apply-join?
- No, doing this is a great way to completely corrupt your installation.
Is it possible to actually export everything from a clustered WebGUI and import it elsewhere?
- Yes, disjoin the cluster and export the configuration. Then rejoin the cluster.
Is it possible to make mass changes without using the GUI? We have a lot of filters and want to create maps using groups of them. It is painfully slow via the map builder - the same holds true for creating a lot of filters.
- This is not a cluster question but the tool you are looking for is called WAAPI.
We are intermittently noticing duplications of pages/page listings in the default folder. How can we prevent that? The likely question is how can we get rid of old data in the cluster database without destroying the cluster? (A little background on this question: We've not been able to determine the circumstances around the occurrence at this time. Is it possible the cluster database is returning outdated information after a restart? We believe the duplicates initially occurred as a result of modification of the pages on the individual nodes - while unjoined, but the cluster still existed. We noticed other changes we'd created weren't showing up, so we dropped the database, then recreated the database and the cluster. Initially all was well. Then suddenly one day the Default folder displays multiple listings of the pages. Looking in the folder under Pages, the old pages were also listed - though they'd been deleted the previous week.)
- I often see people miss configure clustered servers. It sounds like you are one of those people. You should probably open a PMR and ask for some help.
It seems clustering nodes for the Web GUI will result in a lot of downtime if changes need to be made due to the requirement to disjoin, destroy the cluster, the recreate it, so the cluster database is updated. If changes are made on a particular node to the configuration files a lot of manual effort is required replicating the changes on each node.
- Not really if you have configured things properly most of the mentioned operations would not result in downtime.
What are the Best Practices for the clustered WebGUI?