Networker

Re: [Networker] nsrjb slow / lock ups

2004-01-14 08:15:44
Subject: Re: [Networker] nsrjb slow / lock ups
From: Stan Horwitz <stan AT TEMPLE DOT EDU>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Wed, 14 Jan 2004 08:14:53 -0500
Thanks, but this is all well and good and interesting tips, but it
does not apply to the situation I experienced. I went for several
months back and forth with Legato tech support before we  hit upon
a way around the slowness problem that I was experiencing on my
Legato server.

On Wed, 14 Jan 2004, Björn Lundström wrote:

>Hi,
>
>This problem is caused by the RDAC driver that SUN uses with rm6.
>This is also the fact for LSI Logic Storage systems with the RDAC fail-over
>driver installed. Following these instructions often clears the problem.
>
>
>You can actually make it work fast by:
>
>editing :
>
>/kernel/drv/sd.conf
>
>Remove all luns that you do not plan to use in the near future.
>For LSI Logic, also leave The LUN number that you use for managing the
>array.
>
>/kernel/drv/rdnexus.conf
>
>Deleting entries (rows) 10-64 is normally fine.
>
>But there need to be an rdnexus entry for each tru SCSI/FC bus in the system
>so count how many buses you have first.
>
>After this a reboot is required.
>
>After that nsrjb normally works fine , and boot and inquire will run much
>faster.
>
>Best Regards
>Björn Lundström
>
>-----Original Message-----
>From: T. S. Kimball [mailto:tkimball AT BRASS DOT COM]
>Sent: den 14 januari 2004 08:28
>To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
>Subject: Re: [Networker] nsrjb slow / lock ups
>
>
>Actually, the Sun release notes (for us SBU users anyway) mention that
>SBU/Legato has conflict issues with the 'rm6' RAID software used on
>the A1000 and similar units.  inquire will be nearly unresponsive and
>nsrjb operations will take long time.
>
>I actually have an E6000 with 12 CPUs and a little L7 (ADIC Scalor I
>think), take 1 hour to run inventory operations on 7 slots.  rm6 is
>controlling an A1000 and E5300FC (I think) on the same unit.  A
>similar L7 attatched to a 2-CPU Ultra-2 is perfectly fine, and my
>8-drive L700 is also happy with a 4-CPU E450 (both of those have no
>rm6 RAIDs).
>
>
>--TSK
>
>
>On Tue, 13 Jan 2004, Stan Horwitz wrote:
>
>> On Tue, 13 Jan 2004, Olaf Zaplinski wrote:
>>
>>>Hi *,
>>>
>>>when our NW 6.1.4/Linux/Storagetek L700e is busy somehow, it often
>>>ignores tape mount requests and manual tape operations.
>>>
>>>Right now, it wants 5 tapes for backup'ing to pool Default, 4 tapes
>>>for backup'ing to another pool. Out of 5 LTO drives, two are in
>>>state 'ejected'. This is the status for several hours now. Why does
>>>it not mount two volumes in a 'first come, first served' fashion and
>>>starts backup'ing that group?
>>
>> I had this same problem with NetWorker 6.1.3 under Sun Solaris.
>> There really is only one solution. You either have to mount fewer
>> tapes or increase the processing power of your server. The nsrd
>> process (which controls all the activity of the software) chokes
>> when it cannot get enough processing power and the result is the
>> behavior you describe.
>>
>> I resolved the issue by reorganizing my backup schedule and groups
>> so that fewer tapes are requested in a given period of time AND
>> adding three more CPU boards and additional RAM really helped
>> enormously!
>>
>> If you have a tech support agreement with Legato, you should
>> definitely open up a call with them, but odds are, you will get the
>> same response from Legato that I just provided.
>>
>
>
>
>STATEMENT OF CONFIDENTIALITY
>
>The information contained in this electronic message and any attachments
>to this message are intended for the exclusive use of the addressee(s)
>and may contain confidential or privileged information. If you are not
>the intended recipient, please notify SunGard Trading Systems immediately
>at (201) 499-5900 and destroy all copies of this message and any
>attachments.
>
>--
>Note: To sign off this list, send a "signoff networker" command via email
>to listserv AT listmail.temple DOT edu or visit the list's Web site at
>http://listmail.temple.edu/archives/networker.html where you can
>also view and post messages to the list.
>=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
>
>--
>Note: To sign off this list, send a "signoff networker" command via email
>to listserv AT listmail.temple DOT edu or visit the list's Web site at
>http://listmail.temple.edu/archives/networker.html where you can
>also view and post messages to the list.
>=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
>

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

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