Vlad
Cloudstack agent.out or log on SSVM would tell you the last operation
(or set of operations) it performed.
Perhaps you can replay them by hand in shell to find the culprit?
Also, if i read correctly, the NFS is a soft mount, hence you should not
have system lockups. I suspect an underlying problem either in
hypervisor or network stack..
Regards
ilya
On 5/26/15 10:55 AM, Vladislav Nazarenko wrote:
Hi Everybody,
I'm expierencing some performance problems with SSVM.
/usr/local/cloud/systemvm/ssvm-check.sh shows no problems
When the agent is running, I see 100% of io-wait.
secondary storage is mounted, but due to the io-wait, I can not access it
When I stop the agent(/etc/init.d/cloud stop), I can easily access the
nfs mount. I also tried to download images manually (curl), works good.
the physical host hos not IO problems
/etc/exports:
/storage 10.0.40.0/24(rw,async,no_root_squash,no_subtree_check)
Mount in the SSVM:
10.0.40.10:/storage/secondary on
/mnt/SecStorage/0405ba30-38b9-349d-9ea4-15ea09d67e09 type nfs
(rw,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,acdirmin=0,acdirmax=0,soft,proto=tcp,timeo=133,retrans=2147483647,sec=sys,mountaddr=10.0.40.10,mountvers=3,mountport=892,mountproto=tcp,local_lock=none,addr=10.0.40.10)
Cloudtack: 4.4.3 on centos 7
Please advice what to check
Vlad
