Howard,
Your backup strategy will not work correctly.
The fact that you are doing incrementals for each
nodename is causing a serious problems because a restore
scenario requires a Full and ALL incrementals in order
to restore correctly. An Incremental will truncate logs,
and that is where the problems begin.
In your case, your incrementals are spanning multiple nodes.
So, your separate nodes will not contain all of the
logs necessary for a proper recovery of the Incrementals.
If you are sending the backups to different node names,
they will both step on each other and cause serious headaches
when you try to restore.
You have a few options... you could do something like this:
Sun - Full-weekly
Sun - Full-exch
Mon - Incr-exch
Tue - Incr-exch
Wed - Incr-exch
Thu - Incr-exch
Fri - Incr-exch
Sat - Incr-exch
then repeat.
Or you could use a single nodename, and use a COPY type backup
in order to bind those backups to a different management class.
Like this:
Sun - Copy-exch
Sun - Full-exch
Mon - Incr-exch
Tue - Incr-exch
Wed - Incr-exch
Thu - Incr-exch
Fri - Incr-exch
Sat - Incr-exch
In your strategy, you will be able to restore all of Full backups
successfully. But, you will not be able to restore all of the
Full + Incrementals correctly.
Thanks,
Del
----------------------------------------------------
"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU> wrote on 01/08/2010
12:37:41 PM:
> [image removed]
>
> Dissapearing Incrementals
>
> Howard Coles
>
> to:
>
> ADSM-L
>
> 01/08/2010 12:38 PM
>
> Sent by:
>
> "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
>
> Please respond to "ADSM: Dist Stor Manager"
>
> Ladies and Gents, I have a very perplexing problem.
>
>
>
> I have 3 Exchange servers that backup in two ways.
>
> They each have a "-exch" node name for daily backups for DR purposes
> and general purpose restores.
>
> Then they also have a "-weekly" node name for weekly backups for
> archival purposes (yea, I know we have purchased a real email archive
> solution so this will go away soon).
>
>
>
> Now, I backup the -exch nodes full each Sunday, and do incremental
> through the rest of the week, and I do a similar backup with the -weekly
> except it's a full on the first of the month, with each weekly backup
> after that being an incremental.
>
>
>
> Here's where it gets weird. The Incrementals for the -weekly nodes
> appear to be gone.
>
> Running: tdpexcc query TSM * INCR /ALL /TSMOPTF=dsm.weekly.opt
>
> Returns :
>
> ACN5820E There were no Exchange backups found on the Tivoli Storage
> Manager server matching the specified criteria.
>
>
>
> Now, fulls and diffs show up just fine.
>
>
>
> IF I run the same query against the -exch nodes I get the info on all
> the incrementals.
>
>
>
> The same is true in the GUI interfaces as well.
>
>
>
> TSM Server version: 5.5.3 (recently patched) running on AIX 6.1
>
> TSM BA Client version: 5.5.1 (yea, going to update to 5.5.2.7 but not
> until this problem is resolved).
>
> TSM TDP Client version: 5.5.1
>
>
>
> Here's the log entry that shows the Incremental ran:
>
>
>
> 12/12/2009 20:05:15 ============== Log file pruned using log retention
> period of 60 day(s)
>
> 12/12/2009 20:05:15 ============== 22 out of 198 entries pruned from the
> top of this log
>
> 12/12/2009 20:05:16
> ========================================================================
> =
>
> 12/12/2009 20:05:16 Request : Backup
>
> 12/12/2009 20:05:16 SG List : *
>
> 12/12/2009 20:05:16 Backup Type : INCREMENTAL
>
> 12/12/2009 20:05:16 Database Name :
>
> 12/12/2009 20:05:16 Buffers : 3
>
> 12/12/2009 20:05:16 Buffersize : 1024
>
> 12/12/2009 20:05:16 Exchange Server : *******EXCH**
>
> 12/12/2009 20:05:16 TSM Node Name :
>
> 12/12/2009 20:05:16 TSM Options File : dsm.weekly.opt
>
> 12/12/2009 20:05:16 Mount Wait : Yes
>
> 12/12/2009 20:05:16 Quiet : No
>
> 12/12/2009 20:05:16
> ------------------------------------------------------------------------
> -
>
> 12/12/2009 20:07:08 Total storage groups requested for backup: 4
>
> 12/12/2009 20:07:08 Total storage groups backed up: 4
>
> 12/12/2009 20:07:08 Total storage groups expired: 0
>
> 12/12/2009 20:07:08 Total storage groups excluded: 0
>
> 12/12/2009 20:07:08 Throughput rate:
> 17,087.00 Kb/Sec
>
> 12/12/2009 20:07:08 Total bytes transferred:
> 1,924,154,600
>
> 12/12/2009 20:07:08 Elapsed processing time: 109.97
> Secs
>
>
>
> Has anyone else seen this, or heard of this? Folks here are kind of
> panicking because the restore they need is for an exec (CxO).
>
>
>
> I wanted to seek it out here before I opened a ticket with IBM just in
> case someone already resolved this problem.
>
>
>
> See Ya'
>
> Howard Coles Jr.
>
> Sr. Systems Engineer
>
> (615) 296-3416
>
> John 3:16!
>
>
>
>
|
