UPDATE! The community will be go into read-only on April 19, 8am Pacific in preparation for migration on April 21. Read more.
UPDATE! The community will be go into read-only on April 19, 8am Pacific in preparation for migration on April 21.Read more.
Commodore
Commodore
851 views

OES2015, GW, and kworker/flush

Hello all,

I've recently updated to OES 2015 on my Groupwise server (4cpu/16GB VM), and I THINK I'm seeing some NSS issues. I have Groupwise data on an NSS volume, 3 post offices: 2 larger ones (one of which is for students that is quite underutilized) and one tiny one. All together, my entire GW system is about 340GB, 320 of that is in the staff PO.
When the staff PO is loaded, I see the server acting quite sluggish, but can't really identify why.

It's not swapping much (free -m:Swap: 4102 354 3748),
Has nearly non existent IO wait. (vmstat's wa: 1)
Perf shows the following at the top (after doing a perf -g -a record):
+     71.76%          swapper  [kernel.kallsyms]                   native_safe_halt                  ↑
+ 2.20% kworker/1:0 [kernel.kallsyms] _raw_spin_unlock_irqrestore ▮
+ 2.06% kworker/0:1 [kernel.kallsyms] _raw_spin_unlock_irqrestore ▒
+ 1.37% flush-8:0 [kernel.kallsyms] _raw_spin_unlock_irqrestore ▒


But I admittedly don't know how to interpret that.

Aside from the rare times that gwpoa, ndsd or java peak up to the top of the list, 'top' consistently shows a few 'kworker/3:0' or similar threads at the top, consuming anywhere from 10-80 percent of cpu each. Occasionally, "flush-8:0" jumps to the top of the list.
If i unload that POA, it seems to snap right back to life.
Can anyone suggest anything else I can look at to find out what's dogging this server down?

Labels (2)
0 Likes
2 Replies
Commodore
Commodore

Looks like the POA and NSS thoughts might have been a red herring.
I did a "hardware" double check, and I had forgotten to delete the snapshot that I took just before updating to OES2015. The larger the changed disks got, the worse performance got. I guess GW on NSS really doesn't like having a snapshot on it. 😕
I deleted it, and it's back to it's old snappy self.

0 Likes
Knowledge Partner Knowledge Partner
Knowledge Partner

Am 17.11.2015 um 21:46 schrieb adrockk:
>
> Looks like the POA and NSS thoughts might have been a red herring.
> I did a "hardware" double check, and I had forgotten to delete the
> snapshot that I took just before updating to OES2015. The larger the
> changed disks got, the worse performance got. I guess GW on NSS really
> doesn't like having a snapshot on it. 😕
> I deleted it, and it's back to it's old snappy self.
>
>

Thanks for the report back. And yes, snapshots *do* kill I/O
performance, and GW is very sensitive to it.

CU,
--
Massimo Rosen
Novell Knowledge Partner
No emails please!
http://www.cfc-it.de
CU,
--
Massimo Rosen
Micro Focus Knowledge Partner
No emails please!
http://www.cfc-it.de
0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.