Re: Disks take long time to bring online if failover
- From: "George He" <georgeh@xxxxxxxxxxxx>
- Date: Mon, 8 Aug 2005 13:08:33 +1000
Thanks,
No chkDsk_...log found, no logs in the event log, could not find out any
clue in cluster. May try to contact MS.
Regards
George
"John Toner [MVP]" <jtoner@xxxxxxxxxxxxxxxxxxxxx> wrote in message
news:ecXc1xDmFHA.1480@xxxxxxxxxxxxxxxxxxxxxxx
> If it were doing chkdsk, you would see files named
> "ChkDsk_Disk12_Sig00B6147A.log" or similar. Any events in the event log
> while you're bringing the disks online? You also might want to look
> through
> the cluster.log to see if there are any hints while the disk is coming
> online.
>
> You should also check your HBA firmware and drivers to make sure that
> these
> are current. You might want to open a case with MSFT to look into this
> further.
>
> Regards,
> John
>
>
> "George He" <georgeh@xxxxxxxxxxxx> wrote in message
> news:u2CCEh8lFHA.3884@xxxxxxxxxxxxxxxxxxxxxxx
>> Seems no chkdsk run on OS.
>>
>> Here are the files in the Cluster directory, which file should I look
> into?
>>
>> C:\WINDOWS\Cluster>dir /s
>> Volume in drive C has no label.
>> Volume Serial Number is 3CDC-F842
>>
>> Directory of C:\WINDOWS\Cluster
>>
>> 18/07/2005 02:18 PM <DIR> .
>> 18/07/2005 02:18 PM <DIR> ..
>> 24/03/2005 05:57 PM 318,464 cladmwiz.dll
>> 24/03/2005 05:57 PM 873,984 clcfgsrv.dll
>> 24/03/2005 05:57 PM 16,343 clcfgsrv.inf
>> 24/03/2005 05:57 PM 56,832 clnetres.dll
>> 24/03/2005 05:57 PM 52,736 clnetrex.dll
>> 24/03/2005 05:57 PM 176,128 cluadmex.dll
>> 24/03/2005 05:57 PM 443,904 cluadmin.exe
>> 30/03/2003 12:00 AM 52,224 CluAdMMC.dll
>> 01/08/2005 09:19 AM 262,144 CLUSDB
>> 24/03/2005 05:57 PM 476,160 clusres.dll
>> 24/03/2005 05:57 PM 838,656 clussvc.exe
>> 03/08/2005 11:07 AM 4,282,100 cluster.log
>> 01/08/2005 09:17 AM 195,052 cluster.oml
>> 24/03/2005 06:00 PM 41,472 debugex.dll
>> 24/03/2005 06:07 PM 80,384 mqclus.dll
>> 03/06/2005 01:24 PM <DIR> MSCS
>> 24/03/2005 06:13 PM 68,096 resrcmon.exe
>> 30/03/2003 12:00 AM 30,720 VSSTask.dll
>> 30/03/2003 12:00 AM 45,568 VSSTskEx.dll
>> 30/03/2003 12:00 AM 7,680 WSHClus.dll
>> 19 File(s) 8,318,647 bytes
>>
>> Directory of C:\WINDOWS\Cluster\MSCS
>>
>> 03/06/2005 01:24 PM <DIR> .
>> 03/06/2005 01:24 PM <DIR> ..
>> 03/06/2005 01:24 PM 776 quotomb.stn
>> 1 File(s) 776 bytes
>>
>> Total Files Listed:
>> 20 File(s) 8,319,423 bytes
>> 5 Dir(s) 6,218,973,184 bytes free
>>
>> C:\WINDOWS\Cluster>
>>
>> Thanks
>>
>> George
>>
>> "Jim" <Jim@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
>> news:EDD73293-A1D4-41F9-9612-101D36B9071D@xxxxxxxxxxxxxxxx
>> >I saw this thread and we are having the same issue and yes....the O/S is
>> > running chkdsk on the volumes.....
>> >
>> > "John Toner [MVP]" wrote:
>> >
>> >> Is the O/S doing a chkdsk on the volumes before bringing them online?
>> >> Check
>> >> the c:\windows\cluster folder for chkdsk output files
>> >>
>> >> Regards,
>> >> John
>> >>
>> >> "George He" <georgeh@xxxxxxxxxxxx> wrote in message
>> >> news:OrsJnebkFHA.1372@xxxxxxxxxxxxxxxxxxxxxxx
>> >> > To summary:
>> >> >
>> >> > 1. The registry key as you suggested:
>> >> > 2. I could see the physical disk from Powerpath and from Windows
> device
>> >> > manager, so the SAN should be fine
>> >> > 3. The small disk-- qurom disk could bring online very quick
> (something
>> >> less
>> >> > than 30s)
>> >> > 4. The larger cluster disks (each around 2TB) take long time to
>> >> > bring
>> >> > online, it was fine when they were empty (no data)
>> >> >
>> >> >
>> >> > Regards
>> >> >
>> >> > George
>> >> >
>> >> >
>> >> >
>> >> > "John Toner [MVP]" <jtoner@xxxxxxxxxxxxxxxxxxxxx> wrote in message
>> >> > news:eZnQvlTkFHA.2916@xxxxxxxxxxxxxxxxxxxxxxx
>> >> > > Are you using PowerPath on the hosts? If so, try changing the
>> >> > > following
>> >> > > registry key:
>> >> > >
>> >> > >
> HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\EmcpBase\Parameters
>> >> > >
>> >> > > Add/Change the "MpPagingIoTimeout"=dword:00000000
>> >> > >
>> >> > > Regards,
>> >> > > John
>> >> > >
>> >> > > "George He" <georgeh@xxxxxxxxxxxx> wrote in message
>> >> > > news:uEFiIIOkFHA.576@xxxxxxxxxxxxxxxxxxxxxxx
>> >> > >> Hello,
>> >> > >>
>> >> > >> I have a Windows 2003 SP1 two nodes cluster attach to an EMC
>> >> > >> CX500
>> >> > >> SAN.
>> >> > > They
>> >> > >> have 7 physical disks which are presented by SAN. the cluster was
>> >> > >> fine
>> >> > >> before we migrate data to the cluster. e.g. failover without any
>> >> > >> problems.
>> >> > >> Now, with 7TB data on the SAN, if we failover from one node to
>> >> > >> another
>> >> > > node,
>> >> > >> the physical disk resource take around 10 to 15 minutes to bring
>> >> online.
>> >> > >>
>> >> > >> Anyone has any ideas on why it behaviours like this? I suspect it
> is
>> >> > > because
>> >> > >> there are so many data on these disks.
>> >> > >>
>> >> > >>
>> >> > >> Thanks
>> >> > >>
>> >> > >> George
>> >> > >>
>> >> > >>
>> >> > >
>> >> > >
>> >> >
>> >> >
>> >>
>> >>
>> >>
>>
>>
>
>
.
- References:
- Re: Disks take long time to bring online if failover
- From: Jim
- Re: Disks take long time to bring online if failover
- From: George He
- Re: Disks take long time to bring online if failover
- From: John Toner [MVP]
- Re: Disks take long time to bring online if failover
- Prev by Date: Re: Pinging Public Cluster IP is unpredictable
- Next by Date: MS Windows Cluster System for simulation
- Previous by thread: Re: Disks take long time to bring online if failover
- Next by thread: Re: Disks take long time to bring online if failover
- Index(es):
Relevant Pages
|
Loading
