OnDemand User Group

Support Forums => MP Server => Topic started by: jsquizz on September 13, 2017, 01:00:46 PM

Title: arsmaint flag differences
Post by: jsquizz on September 13, 2017, 01:00:46 PM
Greetings folks,

Can anyone give me a high level idea of the differences in the two arsmaint flags..Ive read the IBM documentation and I cant grasp it.

 -d Expire Database
 -i Expire Migrated Imported Database Tables

to me they sound the same.  Thanks
Title: Re: arsmaint flag differences
Post by: Justin Derrick on September 14, 2017, 04:23:59 AM
It's extremely unlikely that you want to use the -i option.

20 years ago when I started with IBM CMOD, storage was so prohibitively expensive that you could 'archive' database tables into TSM.  If someone wanted to do a query from an archived database table, you would pull it out of TSM and reload it, then do the search.  The table would stay in the database until it was expired off with -i.

Today, it doesn't make sense to send the metadata to be archived -- storage is practically free by comparison to the costs 20 years ago.

As a funny analogy, I remember a conversation with our Enterprise Backup team about how to back up IBM CMOD -- they insisted that it wasn't possible to back up 75GB (gigabytes!) on a nightly basis.  I have several 128GB USB sticks within arm's reach here.  :) 
Title: Re: arsmaint flag differences
Post by: jsquizz on September 14, 2017, 06:20:53 AM
Thanks, this has always been a grey area for me. And in discussions with coworkers, both current and past. Some of them with years on me, it was also a grey area for them...arsmaint in general. Since it was usually set a long time ago, and since then people have rotated around and there really was no need to touch what wasn't broken.

So, ARSMAINT will actually expire the database indexes..While TSM will handle the actual deletion of the documents/data.
Title: Re: arsmaint flag differences
Post by: Justin Derrick on September 14, 2017, 07:45:39 AM
Expiration is extremely complicated, and there's different behaviours for different settings of Expiration Types, and how you've configured the policies in TSM.

Essentially, you'll never use -i, because you'll never archive IBM CMOD tables.  Using -d will expire the metadata in the tables, and -c will clear data out of the cache.  Depending on the Expiration Type that's set for the Application Group, IBM CMOD might tell TSM to remove some objects, or it might just leave TSM to do it's own expiration.

-JD.