Geoffrey,

Similar to what we see. In our case just plain old SCSI connected devices so
I can take out all of the various SAN questions from your example. When it
(journaling) works it's great! But we have nights where the backup will
fail. If we stop the TSM Journal Service and restart it the backup will run.
There are times when the Journaling Service won't stop or gets and error
when it is trying to stop. Also have had it stop successfully and then when
trying to restart it will fail stating the service is already active when
based on the Services GUI it isn't running. This is on 5.1.1 NT4 SP6
clients. I haven't taken this to support yet. It's on my list just looking
for time to collect some additional data and open a call!

We have three Intel file servers that are similar in size.
4.5 - 5 million files spread over 20 different drives. Different issue is
that we can only setup the Journal on 10 drives or so. After that we get an
error.
Typically 120-130 GB per night backed up per client.
Journaling on - right around 8 hours - If we could Journal all of the drives
I would expect this time to go down even more.
Journaling off - 12 + hours.

Server - AIX 4.3.3 ML10 / TSM 5.1.1.4

Curt Magura
Lockheed Martin EIS
Orlando, Fla.
321-235-1203


-----Original Message-----
From: Gill, Geoffrey L. [mailto:GEOFFREY.L.GILL AT SAIC DOT COM]
Sent: Tuesday, November 12, 2002 5:32 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Journaling


This journaling thing, when working, is definitely a help. I must say though
that in this specific case it has been mighty strange. I set this up back in
October and it did not kick in on this SAN disk with the 4+ million files
right away. But there was an entry on 11/2 that pointed to a completion, I
could tell from the amount of files inspected and the entry "using journal
on g$". Then all of a sudden for about a week it stopped.

I called support, sent all the files and was told it doesn't work with SAN
attached storage. After some emails from all of you I decided to re-open
this ticket, which I did. In the mean time I've been looking at the logs and
for the past 2 backups it's now working again. Don't ask me how.

I'm going to respond to support accordingly when they call me, and send them
all the files I have, again. I'd like to see if they can find a reason for
the 6 day break if possible. I'd also like to know, when in the log I sent
them it clearly shows it was working on the 2nd, why they'd tell me it's not
supported on SAN attached disk.

Geoff Gill
TSM Administrator
NT Systems Support Engineer
SAIC
E-Mail:   gillg AT saic DOT com
Phone:  (858) 826-4062
Pager:   (877) 905-7154


> -----Original Message-----
> From: Chris Murphy [mailto:cmurphy AT IDL.STATE.ID DOT US]
> Sent: Tuesday, November 12, 2002 1:57 PM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: Journaling
>
> Geoff,
>
> Clients: NT4 SP6 and Win2k SP2/3 - TSM client 5.1.1.0
> TSM Server: Win2k SP3 - V5.1.1.0
>
> FWIW, we currently have 3 servers that fit within Zlatko's #4 category.
> Between all the servers there is about 4+ million files and about 1TB of
> data on an HP disk array.  Journaling  is on and works great.  We average
> about 1-5% changed files each night and backups typically take about 15-45
> minutes for all 3 servers!
>
> And I believe Zlatko is correct on his categorizing.  Windows "sees" disk
of
> type #1,2,3,4&5 as local.  (#5 is a bit of a special case however...)  It
> cannot tell the difference between a DASD SCSI/RAID disk, and a
SAN-attached
> FC array.  Windows "speaks" SCSI to the SAN disks the same as it does to
> DASD's and the HBA simply handles the FC encapsulation and transport
(since
> FC really just does exactly that: encapsulates the SCSI commands).  OS
> has/needs no knowledge of it.  I would call up support again, as it SHOULD
> work.
>
> My two bits...
>
> Chris Murphy
> IT Network Analyst
> ID Dept. of Lands
> Office: (208) 334-0293
> cmurphy AT idl.state.id DOT us