Last night (first since the upgrade from 7.3.5.1P2 to 8.1RC2) we experienced much higher load on our 3270 cluster and some VM timeouts coinciding with our dedup schedule (11pm -2am)
Anyone heard if dedup changed in 8.1?
I am disabling the regular dedup schedule until this is sorted out
thanks
Fletcher
Id grab a lot more data, than CPU use...which indicates very little on it's own.
I agree you had a problem, but CPU % wasn't the cause.
Say hi to Vivian for me! :)
On Thu, Feb 9, 2012 at 9:40 AM, Fletcher Cocquyt fcocquyt@stanford.eduwrote:
Last night (first since the upgrade from 7.3.5.1P2 to 8.1RC2) we experienced much higher load on our 3270 cluster and some VM timeouts coinciding with our dedup schedule (11pm -2am)
Anyone heard if dedup changed in 8.1?
I am disabling the regular dedup schedule until this is sorted out
thanks
Fletcher
Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
Just thought I would put this out there..
Couple days ago, started seeing these errors in my message file.. Opened the case with support, and they said its "normal" and is going to disk for the data.. Has anyone seen this in their messages file before? I am waiting to see if it's a pattern, but has not gone away, and has happened the last couple days.
: extCache.io.readError:error]: WAFL external cache I/O read error: FAL error: fal_response_get_status() failed , code 2.
This filer is a 6240, PAMII, DOT 7.3.6.p2 running CIFS/FC, etc. No VMWare, etc. Couple heavy SQL servers on it.
I guess I could press and ask for a new PAM card (which was replaced about 6 months ago). Or take it down for diags..
Has anyone seem this before?
Thanks
Steve
Ya. Just a bad block or something. Pam is not your primary block home anyways. Not worth downtime.
.
On Feb 9, 2012, at 10:04 AM, "Klise, Steve" klises@sutterhealth.org wrote:
Just thought I would put this out there..
Couple days ago, started seeing these errors in my message file.. Opened the case with support, and they said its “normal” and is going to disk for the data.. Has anyone seen this in their messages file before? I am waiting to see if it’s a pattern, but has not gone away, and has happened the last couple days.
: extCache.io.readError:error]: WAFL external cache I/O read error: FAL error: fal_response_get_status() failed , code 2.
This filer is a 6240, PAMII, DOT 7.3.6.p2 running CIFS/FC, etc. No VMWare, etc. Couple heavy SQL servers on it.
I guess I could press and ask for a new PAM card (which was replaced about 6 months ago). Or take it down for diags..
Has anyone seem this before?
Thanks
Steve
Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
Day2 after the 8.1RC2 upgrade - All quiet on the (disabled) dedup front - will need to engage the lead engineers on this one since we like the dedup savings, but not the 100's of VMs starving for IO.
Fletcher Cocquyt Principal Engineer Information Resources and Technology (IRT) Stanford University School of Medicine
Email: fcocquyt@stanford.edu Phone: (650) 724-7485
On Feb 9, 2012, at 9:40 AM, Fletcher Cocquyt wrote:
Last night (first since the upgrade from 7.3.5.1P2 to 8.1RC2) we experienced much higher load on our 3270 cluster and some VM timeouts coinciding with our dedup schedule (11pm -2am)
Anyone heard if dedup changed in 8.1?
I am disabling the regular dedup schedule until this is sorted out
<PastedGraphic-13.png>
thanks
Fletcher
Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
Day2 after the 8.1RC2 upgrade - All quiet on the (disabled) dedup front - will need to engage the lead engineers on this one since we like the dedup savings, but not the 100's of VMs starving for IO.
After a major upgrade, ONTAP usually runs a background job to upgrade the WAFL filesystem. Maybe that is what caused your load problems. Presumably that has finished, so perhaps you can enable dedupe again.
Last night (first since the upgrade from 7.3.5.1P2 to 8.1RC2) we experienced much higher load on our 3270 cluster and some VM timeouts coinciding with our dedup schedule (11pm -2am)
Anyone heard if dedup changed in 8.1?
I am disabling the regular dedup schedule until this is sorted out
<PastedGraphic-13.png>
thanks
Fletcher
Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
Steve Losen scl@virginia.edu phone: 434-924-0640
University of Virginia ITC Unix Support
-
Fletcher Cocquyt -
Jeff Mohler -
JM
-
Klise, Steve -
Steve Losen