Index Management for Cisco Tidal Enterprise Scheduler Client Manager Database

[Shyam Sundar Srinivasan]

In our TES 6.x architecture, Client Manager plays a key role in serving the requests coming in from Web Client, Transporter, CLI, iPhone client and REST APIs. Hence it is all the more important to ensure that the Client Manager is sized and tuned properly for an optimal performance.

An important step in maintaining a healthy Client Manager is to ensure that the cache database (TESCache) is well maintained. As like maintaining any other database, we recommend our customers to involve their DBAs every once in a while to ensure that the CM cache DB is maintained properly.

Some of the common DB maintenance activities include:

1. Index management
2. Monitoring full table scans
3. Monitoring CPU, Memory and Disk usage
4. Monitoring database fragmentation

In this post I’m attaching a document from our engineering group that captures recommendations on managing indexes in your TESCache to boost your web client performance. Hope you find this useful.

Thanks,

Shyam Sundar Srinivasan

Product Manager

Cisco Tidal Enterprise Scheduler

[Brian Dickens]

Shyam,

What you say makes sense and I agree. With respect to the article you attach though, I'm curious. It reads like it's a test plan to see if the changes suggested have an improvement. It doesn't read like a suggestion for all users to implement in their Client Manager cache databases. Are you suggesting that we make these changes? If this improves performance, why not implement it via a hotfix?

It's been my experience that when there's an issue with the Client Manager, one of the first steps toward fixing the issue is to drop all the tables and let the CM recreate them.

Thanks.

[Joe Fletcher]

Interestingly enough we've seen problems with the Client Manager database essentially not creating indexes when it should. Its an on-going issue but I'll read through your doc in case there is anything to suggest a fix.