ADSM-L

Re: Where does the data go

2002-10-17 22:49:01
Subject: Re: Where does the data go
From: "Seay, Paul" <seay_pd AT NAPTHEON DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 17 Oct 2002 22:47:23 -0400
Can you state what is going on!  Migration has nothing to do with copy
pools.  Each backup object can have only one primary sequential pool
associated with it unless the cache on disk is used, but it still gets
migrated.  These objects have additional objects associated that are the
copy storage pool objects.

This is an example of the show bfo (Bit File Object) for a backups table
entry.  You will notice this is what connects the backups table to the
contents table.

**Sub-bitfile 0.203027760 is stored in the following aggregate(s)
  Super-bitfile: 0.203027760, Offset: 0.0, Length 0.646

**Archival Bitfile Entry
  Bitfile Type: PRIMARY  Storage Format: 9
  Bitfile Size: 0.702823  Number of Segments: 1
  Storage Pool ID: 10  Volume ID: 8513  Volume Name: 421705
**Archival Bitfile Entry
  Bitfile Type: COPY  Storage Format: 9
  Bitfile Size: 0.702823  Number of Segments: 1
  Storage Pool ID: -3  Volume ID: 8493  Volume Name: 400586
**Archival Bitfile Entry
  Bitfile Type: COPY  Storage Format: 9
  Bitfile Size: 0.702823  Number of Segments: 1
  Storage Pool ID: -13  Volume ID: 8538  Volume Name: 423547

Typically, if collocation is set to NO migration will fill up a cartridge
unless you are specifying many more than process maxpr=x.

If you have collocation turned on.  All bets are off.  Migration will make
every attempt to collocate the data for a node/filespace.  This usually
means at least one tape per.



Paul D. Seay, Jr.
Technical Specialist
Naptheon Inc.
757-688-8180


-----Original Message-----
From: Sam Haziprodromu [mailto:sam.hazi AT IRWINTOY DOT COM]
Sent: Thursday, October 17, 2002 3:41 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Where does the data go


Hi Tsmers I have a long winded question.When migration runs and forces the
data to be backed to the primary tape pool, why doesn't TSM just pick one
cartridge and fill it up and only then seek another cartridge.In other words
if this is how it really worked then you could get by with one bay in the
LTO. Is the fact that a copy pool cartridge is also sent offsite have
anything to do with my theory.If anyone has any thoughts on this i would
appreciate it if you could please set them down on paper.

                     TIA

                   Sam Haziprodromu  I.T.

<Prev in Thread] Current Thread [Next in Thread>