ADSM-L

Re: [ADSM-L] EXPORTING clients

2011-07-10 11:44:54
Subject: Re: [ADSM-L] EXPORTING clients
From: Zoltan Forray <zforray AT VCU DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Sun, 10 Jul 2011 11:40:17 -0400
What V/R/M/OS are you running?

"Hughes, Timothy" <Timothy.Hughes AT OIT.STATE.NJ DOT US> wrote:

Hi Zoltan,

Thanks...I will take a look at those there on the IBM support site I assume 
correct? I tried looking for something about this issue before but could not 
find anything. I will try again must have missed something.

Regards

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of 
Zoltan Forray
Sent: Saturday, July 09, 2011 11:43 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: EXPORTING clients

IIRC, there are some server patches related to exports causing problems.

"Hughes, Timothy" <Timothy.Hughes AT OIT.STATE.NJ DOT US> wrote:

I think you may be correct, I don't see those input wait times most of the time 
only once or twice since doing the exports. Still our main issue seems to be 
the suspend export command seems to hang and/or cause the destination TSM 
Server to crash. Also running more than one export seems to cause a hang 
sometimes currently running only 1 export (frustrating) cause this is going to 
take forever doing these exports.

regards

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of 
Paul Zarnowski
Sent: Friday, July 08, 2011 2:29 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: EXPORTING clients

Richard,

Here is something to think about:

What Timothy is doing is essentially tape-to-tape data movement for one node, 
with that node's data likely scattered to some degree across the input tape. 
This means that there will be some tape repositioning needed to read the input 
tape. If the output was to disk, it is likely that the input tape could stay 
pretty much in streaming mode. However, with the output being directed to tape, 
IMHO it is much more likely that between the input tape repositions and not 
being able to keep the output tape in streaming mode, the combination will 
likely result in an inordinate amount of backhitching. I don't think this is a 
hardware error, so much as a tape technology limitation. We get around this by 
using serial disk as an interim step, which allows the whole process to run 
more quickly (because there is less overall backhitching).

..Paul


At 08:34 AM 7/8/2011, Richard Sims wrote:
>On Jul 8, 2011, at 7:57 AM, Hughes, Timothy wrote:
>
>> Yes, We are doing Server-to-tape exports. I did notice a couple times long 
>> input tape mounts on the destination TSM SERVER one was around 6,000 seconds 
>> another around 14,000 seconds and another 17,000 seconds.
>
>Export/Import operations are, per the Admin Guide manual topic "Preemption of 
>client or server operations", high priority, which would pre-empt lower 
>priority operations in order to get a tape mount started. If prompt allocation 
>of tape drives is being reflected in your Activity Log, but then tape mount 
>and positioning is taking an inordinate amount of time, that would suggest 
>hardware issues with tapes or drives. If non-Export operations do not exhibit 
>such delays, then something else is going on, where analysis of the Activity 
>Log and operating system logs may reveal factors. If no cause is evident, 
>contacting TSM Support would be in order.
>
> Richard Sims


--
Paul Zarnowski Ph: 607-255-4757
Manager, Storage Services Fx: 607-255-8521
719 Rhodes Hall, Ithaca, NY 14853-3801 Em: psz1 AT cornell DOT edu

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