High system load average after NCS OES/Linux cluster install (repost from n.s.o-e-s.linux.install)

Hi folks,

I'm on about my 6th attempt at installing an OES/Linux SP1 cluster, and
i've managed to overcome most of the issues, with the exception of this
one: as soon as i configure and start NCS, the load average of my system
goes from less than 1 to over 6 and sits there constantly.

The CPU is 95-99% idle, so it seems that the system is creating and
destroying processes constantly.  I've disabled a couple of the more
probable processes (ones that occasionally appear at the top of the
utilization list in top): ndsd and owcimomd, but neither of these make
any difference.  As soon as i 'rmmod cma' (the cluster management kernel
module?), the load average drops back (to about 4 instead of 6).

The system doesn't appear to be performing poorly (although it's hard to
tell without a production load on it).  My hardware is:
- EMC AX100i disk array
- Dell PE SC1425 server, 1 x 3 GHz Xeon EM64T HT, 2 Gb RAM
- HP DL140 G2 server, 1 x 2.8 GHz Xeon EM64T HT, 2 Gb RAM
- HP ProCurve 5308xl switches for both front end and iSCSI VLANs

Questions:
- Has anyone else seen this?
- Is it expected/normal?
- Is there anything i can do about it?
- Could constant writes to the SBD which require interaction with a
userland daemon (i.e. iSCSI) cause this?
- I'm running iSCSI over standard NICs, and my SBD is on an iSCSI
volume.  Assuming an affirmative answer to the above question, should i
consider adding TOE NICs?

Thanks in advance
Paul
0
Paul
12/23/2005 1:15:46 AM
novell.cluster-services 3416 articles. 0 followers. Follow

2 Replies
972 Views

Similar Articles

[PageSpeed] 24

Paul,

It appears that in the past few days you have not received a response to your 
posting.  That concerns us, and has triggered this automated reply.

Has your problem been resolved? If not, you might try one of the following options:
 
- Do a search of our knowledgebase at http://support.novell.com/search/kb_index.jsp 
- Check all of the other support tools and options available at 
http://support.novell.com.
- You could also try posting your message again. Make sure it is posted in the 
correct newsgroup. (http://support.novell.com/forums)

Be sure to read the forum FAQ about what to expect in the way of responses:  
http://support.novell.com/forums/faq_general.html

If this is a reply to a duplicate posting, please ignore and accept our apologies 
and rest assured we will issue a stern reprimand to our posting bot.

Good luck!

Your Novell Product Support Forums Team
http://support.novell.com/forums/


0
Automatic
12/27/2005 9:29:52 AM
Hey Paul,

I dont think that it is anything to do with your hardware config because 
I have the same problem.. I have not looked into this indept yet but 
should be able to sometime next week if all goes well... If you found the 
answer please post it...

Thanks!

-scz


> Hi folks,
> 
> I'm on about my 6th attempt at installing an OES/Linux SP1 cluster, and
> i've managed to overcome most of the issues, with the exception of this
> one: as soon as i configure and start NCS, the load average of my system
> goes from less than 1 to over 6 and sits there constantly.
> 
> The CPU is 95-99% idle, so it seems that the system is creating and
> destroying processes constantly.  I've disabled a couple of the more
> probable processes (ones that occasionally appear at the top of the
> utilization list in top): ndsd and owcimomd, but neither of these make
> any difference.  As soon as i 'rmmod cma' (the cluster management kernel
> module?), the load average drops back (to about 4 instead of 6).
> 
> The system doesn't appear to be performing poorly (although it's hard to
> tell without a production load on it).  My hardware is:
> - EMC AX100i disk array
> - Dell PE SC1425 server, 1 x 3 GHz Xeon EM64T HT, 2 Gb RAM
> - HP DL140 G2 server, 1 x 2.8 GHz Xeon EM64T HT, 2 Gb RAM
> - HP ProCurve 5308xl switches for both front end and iSCSI VLANs
> 
> Questions:
> - Has anyone else seen this?
> - Is it expected/normal?
> - Is there anything i can do about it?
> - Could constant writes to the SBD which require interaction with a
> userland daemon (i.e. iSCSI) cause this?
> - I'm running iSCSI over standard NICs, and my SBD is on an iSCSI
> volume.  Assuming an affirmative answer to the above question, should i
> consider adding TOE NICs?
> 
> Thanks in advance
> Paul

0
zentz
1/6/2006 3:18:04 PM
Reply:

Similar Artilces:

Install 2 nodes cluster : OES Linux and OES Netware
I would like to build a cluster with our Netware OES files server with a new OES Linux server. I seems like there is no installation path when you want to create a brand new cluster with OES Netware and OES Linux. I tried with the nwdeploy but I can't select my OES Linux as the second server and the tool won't let me install a cluster with one server (well that make sense). Anyone know the a good installation path for a mixed two node cluster ? Thanks, Louis-Steve Desjardins Salut The Netware/Linux mixed cluster are intended for migration path. If starting a n...

High system load average after NCS install
Hi folks, I'm on about my 6th attempt at installing an OES/Linux SP1 cluster, and i've managed to overcome most of the issues, with the exception of this one: as soon as i configure and start NCS, the load average of my system goes from less than 1 to over 6 and sits there constantly. The CPU is 95-99% idle, so it seems that the system is creating and destroying processes constantly. I've disabled a couple of the more probable processes (ones that occasionally appear at the top of the utilization list in top): ndsd and owcimomd, but neither of these make any differenc...

OES/Linux SP1 fresh install cluster plugins not working
Hi all, On my squeaky-clean, newly-installed Linux cluster, iManager doesn't work. When i log in as a collection owner, the front screen says "Some of the roles and tasks are not available." When i show details it says: Cluster Manager Clusters Plug-in is not installed. Cluster Event Log Clusters Plug-in is not installed. Cluster Options Clusters Plug-in is not installed. Yet when i go into the iManager configuration and select installed modules, it shows up: Cluster Services 3.0.5.20050812 Cluster Services Management Plugin I found a log file do...

Install gw6.5 on to a cluster or install first and then cluster
Hiya. We have a nw6 sp1 server running as our gw 5.5.5 post office. We will be upgrading to 6.5 very soon. And also installing a SAN and clusters. Is it better to install the SAN and cluster the server that the post office is running on (but probably not cluster the 5.5.5. PO) first and then upgrade to 6.5 or the other way around? Thanks Marcus Since I don't know much about GW 5.5.5 I'd recommend updating GW first to 6.5. There is a switch in the POA settingsfile that tells the POA that it is on a cluster. So I think that it's wise to have GW6.5 fi...

s l o w c o n e c t i o n o n l a p t o p
Now to figure out why the laptop's connection is so much slower than my desktop. Using the same router to the same cable modem to the same cable connection. Desktop bandwidth is measured at around 6.5Mb/s sustained, laptop tops out at about 1.8Mb/s. The NIC in the laptop is rated as a 10/100 ethernet card (Xircom), and Win2K claims a connect speed of 100Mb/s. Drivers for the card are current as of 2002 (I doubt an updated driver would effect this as much as I want anyway). :) Now maybe it was always this slow but I never knew it because my connection back when I was using it ...

Reinstall Cluster Services on Linux OES
Hi, we have two linux OES Sp2 and want to delete and reinstall the Clustering services. Has anyone a documentation how to delete and reinstall the cluster services ??? thx Stefan On Thu, 31 Aug 2006 09:38:57 +0000, sdoern wrote: > Hi, > > we have two linux OES Sp2 and want to delete and reinstall the Clustering > services. > > Has anyone a documentation how to delete and reinstall the cluster > services ??? Are you trying to create a new cluster? If so, do cluster down on one of the nodes, blow away the sbd and then just re-run the yast ncs uti...

NCS OES Linux Load Average
We've installed NCS on a 2 Node OES / Lin cluster. Everything seems to be working correctly, but the Unix load average on both boxes now sits @ 6.00 all the time. This doesn't seem to impact the performance of the boxes in any way, but it's driving our Unix admin crazy. Has anyone else experienced this? Chris Sarrett Netowork Analyst Linn Benton Lincoln ESD Chris.Sarrett@lblesd.k12.or.us On Thu, 05 Jul 2007 23:47:17 +0000, Chris Sarrett wrote: > We've installed NCS on a 2 Node OES / Lin cluster. Everything seems to be > working correctly, but the Un...

Install Netware services on Linux OES
Hi I have been running a Novell network in the office for many years. Server is now a NetWare 6.5 SP4 box (OES). I have also built and have in the tree an OES Linux (SLES9 SP2) box. eDir is installed on the server and both it and the NetWare box are "talking". Server is running LVMS so that NSS partitions can be installed. Server is RAID5 with hardware RAID supported natively by OES/SLES9 SLP is installed on the NetWare box. Workstations are SLED10 and Windows XP, all using the Novell client. I have printed all the documentation I can find but don't have a log...

Where's the installation instructions for Linux???
Name: Anonymous Product: Firefox Summary: Where's the installation instructions for Linux??? Comments: When I download the linux .tar.gz for Firefox 2.0.0.9 from its webpage (see below), there are no installation instructions on the webpage. There are also no installation instructions inside the .tar.gz file. Without installation instructions, novices will give up (and return to Windows and IE, heaven forbid), and seasoned users may still have to search the web to find out how to install Firefox. You could improve this by adding a link on this webpage pointing to some in...

cluster services not starting after upgrade from oes1 linux to oes2 linux
--____DWPVWLNUAHJPQYGTSERX____ Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: base64 Content-Disposition: inline; modification-date="Sun, 19 Jan 2008 11:15:45 +1100" SGkgQWxsLA0KDQpKdXN0IHN0YXJ0ZWQgbXkgZmlyc3QgdXBncmFkZSBvZiBhbiBPRVMxIExpbnV4 IGNsdXN0ZXIgdG8gT0VTMiwgb24gdGhlIGZpcnN0IG5vZGUgSSBhbSB1cGdyYWRpbmcgaXQgd2Vu dCB0aHJ1IHRoZSB1cGdyYWRlIHByb2Nlc3Mgd2l0aG91dCBhbnkgbWFqb3IgZHJhbWFzLCBidXQg dXBvbiBib290aW5nIHRvIE9FUzIgdGhlIHNlcnZlciB3aWxsIG5vdCBzdGFydCBjbHVzdGVyIHNl cnZpY2VzLiBJIHJlY2VpdmUgdGhlIGZvbGxvd2luZyBtZXNzYWdlOi0NCg0Kb2Vzbm9kZTE6fiAj IC9l...

More on New OES Linux Cluster from NW 6.5 Not Clustered
Hello again, As previously stated, we will be creating a clustered environment on OES Linux with a fiber channel SAN. We currently have a non-clustered environment on NW 6.5 with a fiber channel SAN. Clustering aside, as I understand it, it is possible to dismount a volume mounted on a NW 6.5 server and mount it on an OES Linux server. Assuming this is correct, does this apply to clusters as well, where we could dismount a volume mounted on a NW 6.5 server, mount it on the OES Linux server, mark the device Shareable for Clustering, cluster enable the Pool and Volume, crea...

New OES Linux Cluster from NW 6.5 Not Clustered
Hello, We will be creating a clustered environment on OES Linux with a fiber channel SAN. We currently have a non-clustered environment on NW 6.5 with a fiber channel SAN. First question (of likely many) pertains to the special cluster partition. According to the admin guide the NCS installation allocates on cylinder on one drive of the shared disk system for the special cluster partition. Is this a matter of creating a separate LUN for the special cluster partition? On our SAN we've got a number of Raid 5 arrays. Each array has a number of LUNS, and each LUN is de...

It's Linux !? No, it's BSD ! No, no, it's...
kFreeBSD http://wiki.debian.org/Debian_GNU/kFreeBSD a Debian distro using the FreeBSD kernel ! ObiWan wrote: > kFreeBSD > > http://wiki.debian.org/Debian_GNU/kFreeBSD > > a Debian distro using the FreeBSD kernel ! Why use a BSD kernel? I hope the reason is more than, just because they can. http://wiki.debian.org/Debian_GNU/kFreeBSD_why It's easy to understand why the would want to use Debian packages. :) -- Jimmy Johnson Registered Linux User #380263 >> kFreeBSD >> >> http://wiki.debian.org/Debian_GNU/kFreeBSD >> ...

Install in ZCM10 in running oes2 linux cluster
Hallo we have running 3 node oes2 linux cluster with zenworks 7 Desktop Management and Software Packaging, File and Print, etc . Now we want upgrade and migrate to zcm10 in this cluster and would have same functions. Make it sense, is it possible to do that in oes linux cluster? I couldn´t find any documentation about ZCM 10 an does linux clustering. regards Steve Stefan Grzenkowski wrote on Monday, Februar 18th 2008 18:17 in novell.support.zenworks.configuration-management.10.server-install: (...) > we have running 3 node oes2 linux cluster with zenworks 7 Deskt...

Web resources about - High system load average after NCS OES/Linux cluster install (repost from n.s.o-e-s.linux.install) - novell.cluster-services

Resources last updated: 11/22/2015 7:15:51 PM