What sort of tape drives/library is it? It isn't some sort of de-dup'ed
library (non-IBM) is it?
regards,
Mark
From: Niklas Lundström <niklas.lundstrom AT SWEDBANK DOT SE>
To: ADSM-L AT vm.marist DOT edu
Date: 07/09/2010 04:45 PM
Subject: SV: [ADSM-L] SV: [ADSM-L] 50Mb restored in 4.18h....
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT vm.marist DOT edu>
Well I started the restore and went home, so I don't know how many mounts
it did, but the data was on 3 tapes
09/04/2010 11:34:41 ANR1182I Removable volume 006811 is required for a
restore
request from session 2060. (SESSION: 2060)
09/04/2010 11:34:41 ANR1182I Removable volume 005763 is required for a
restore
request from session 2060. (SESSION: 2060)
09/04/2010 11:34:41 ANR1182I Removable volume 005902 is required for a
restore
request from session 2060. (SESSION: 2060)
09/04/2010 11:34:42 ANR1183I Initial determination of removable volumes
required for a restore request from session 2060 is
complete. Additional volumes may still be required.
(SESSION: 2060)
And it looks like it only mounted the tapes once, except for 006811, which
got mounted 2 times:
08:40:39 ADSM4 : q act begind=09/04/2010 endd=09/04/2010 begint=11:00:21
endt=16:00:21 msgno=0510 or=All
Date/Time Message
--------------------
----------------------------------------------------------
09/04/2010 12:07:32 ANR0510I Session 2060 opened input volume 006811.
(SESSION: 2060)
09/04/2010 13:15:42 ANR0510I Session 2103 opened input volume 005763.
(SESSION: 2103)
09/04/2010 15:53:15 ANR0510I Session 2103 opened input volume 005902.
(SESSION: 2103)
09/04/2010 15:54:56 ANR0510I Session 2103 opened input volume 006811.
(SESSION: 2103)
During the restore, the "original" server did a backup, does my restore
somehow collide with the running backup?
09/04/2010 11:37:06 ANR1639I Attributes changed for node FSPAS01FS039: TCP
Name from FSPAS01FS037 to WS25658, TCP Address from
10.190.100.145 to 10.129.22.68, GUID from
c6.05.02.21.fa.22.11.da.85.a0.00.16.35.7c.81.8a to
5e.dd.f5.91.84.8e.11.dd.8d.d6.00.1e.0b.64.bf.1d.
(SESSION: 2103)
FS039 is the fileserver, and my ws is WS25658
This is the results from the actlog:
09/04/2010 15:56:06 ANR0514I Session 2103 closed volume 006811. (SESSION:
2103)
09/04/2010 15:56:09 ANE4955I (Session: 2103, Node: FSPAS01FS039) Total
number
of objects restored: 114 (SESSION: 2103)
09/04/2010 15:56:09 ANE4959I (Session: 2103, Node: FSPAS01FS039) Total
number
of objects failed: 0 (SESSION: 2103)
09/04/2010 15:56:09 ANE4961I (Session: 2103, Node: FSPAS01FS039) Total
number
of bytes transferred: 50.99 MB (SESSION: 2103)
09/04/2010 15:56:09 ANE4963I (Session: 2103, Node: FSPAS01FS039) Data
transfer time: 9,426.15 sec (SESSION:
2103)
-----Ursprungligt meddelande-----
Från: Stefan Folkerts [mailto:stefan.folkerts AT itaa DOT nl]
Skickat: den 7 september 2010 08:30
Till: Niklas Lundström
Ämne: RE: [ADSM-L] SV: [ADSM-L] 50Mb restored in 4.18h....
How many tape mounts did you see during the restore?
-----Original Message-----
From: Niklas Lundström
Sent: 07-09-2010, 08:03
To: ADSM-L AT VM.MARIST DOT EDU
Subject: [ADSM-L] SV: [ADSM-L] 50Mb restored in 4.18h....
I actually restored the files to my workstation, and the network settings
are fine
Duplex mismatch does not cause the speed to go down to modem-speed from
1995....
-----Ursprungligt meddelande-----
Från: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] För Remco
Post
Skickat: den 7 september 2010 07:40
Till: ADSM-L AT VM.MARIST DOT EDU
Ämne: Re: [ADSM-L] 50Mb restored in 4.18h....
On 7 sep 2010, at 07:02, Niklas Lundström wrote:
> Hello
>
> I just did a restore of 50Mb and it took 4h 18min
> If it had been 50Gb I wouldn't complain
>
> Total number of objects restored: 114
> Total number of objects failed: 0
> Total number of bytes transferred: 50.99 MB
> Data transfer time: 9,426.15 sec
> Network data transfer rate: 5.53 KB/sec <--------------
> Aggregate data transfer rate: 3.36 KB/sec
> Elapsed processing time: 04:18:55
>
> The server has a lot of files, almost 7 million including the copypool.
The TSM DB is really big, 180Gb and 94% full.
> TSM Server 5.5.4.
>
> Does anyone have a clue of why the restore took so long?
>
your network sucks? (to put it mildly :) ) I'd check to see if there is any
mismatch in duplex settings anywhere. Remember, auto-negotiation on the
host, and a fixed 100 full-duplex setting on the switch will lead to a host
at half-duplex and tons of collisions.
> It found what volumes to mount fast, but then the client was "Waiting for
files from the server....."
>
> People here are complaining about TSM and that the restores takes long
time
> and it's hard to defend it when the restore takes so long...
>
--
Met vriendelijke groeten/Kind regards,
Remco Post
|