Hi, I believe the EMC VNX iSCSI and NFS traffic are not on the same network interface. I'd check the bottleneck between the hypervisor and the storage, you can use 'iperf' [1] for that. And also, try to check the connectivity to the VNX's NFS ports while the VM is paused while loading data on the VM NFS based disk.
[1] https://iperf.fr/iperf-download.php On Wed, Aug 22, 2018 at 5:01 PM, <jeanbapti...@nfrance.com> wrote: > Hello, > > I'm facing a strange issue on my OVirt Dev pool. > > Indeed, when I create high disk load a VM (kickstart installation or > iozone test for example) , VM is paused due to storage I/O error. > Problem is 100% reproducible, and is located only on NFS (v3 and v4) on my > EMC VNXe3200 NAS's (I have a 10TB and a 20TB NAS) > I have done test (simple iozone -a) with VM 1 vCPU / 2GB RAM and 2 disk > (1*20GB + 1*10GB). Both VMs disk are placed into the same SAN / NAS for > each test. Results are: > - EMC VNXe3200 (10TB) NFSv3 => VM stopped after 10- 30s iozone lauch > - EMC VNXe3200 (20TB) NFSv3 => VM stopped after 10- 30s iozone lauch > - EMC VNXe3200 (10TB) ISCSI => No problem, iozone test finish, and > performance are "standard" regarding load of the VNXe (60MB/s sequential > Write for info) > - EMC VNXe3200 (20TB) ISCSI=> No problem, iozone test finish, and > performance are "standard"regarding load of the VNXe (40-60MB/S sequential > Write for info) > - NETAPP FAS2240 NFSv3 => No problem, iozone test finish, and performance > are good (100MB/s sequential Write for info) > - Freebsd10 NAS NFSv3 => No problem, iozone test finish, and performance > are good regarding NAS conf (80MB/s sequential Write for info) > > I can't explain why I have an issue on NFS and I have not issue on ISCSI > (on the same EMC VNxe3200...). > NFS default params are keeped when storage added to > datacenter:(rw,relatime,vers=4.0,rsize=131072,wsize=131072, > namlen=255,soft,nosharecache,proto=tcp,port=0,timeo=600, > retrans=6,sec=sys,clientaddr=XXXXXXX,local_lock=none,addr=XXXXXX) > (rw,relatime,vers=3,rsize=131072,wsize=131072,namlen= > 255,soft,nolock,nosharecache,proto=tcp,timeo=600,retrans=6, > sec=sys,mountaddr=XXXXXX,mountvers=3,mountport=1234, > mountproto=udp,local_lock=all,addr=XXXXXX) > > Debug logs on host does not help me a lot: > 2018-08-22 15:36:13,883+0200 INFO (periodic/22) [vdsm.api] START > multipath_health() from=internal, task_id=53da2eca-eb66-400c-8367-ab62cedc5dc1 > (api:46) > 2018-08-22 15:36:13,883+0200 INFO (periodic/22) [vdsm.api] FINISH > multipath_health return={} from=internal, > task_id=53da2eca-eb66-400c-8367-ab62cedc5dc1 > (api:52) > 2018-08-22 15:36:15,161+0200 INFO (libvirt/events) [virt.vm] > (vmId='b139a9b9-16bc-40ee-ba84-d1d59e5ce17a') abnormal vm stop device > ua-179375b0-0a18-4fcb-a884-4aeb1c > 8fed97 error eother (vm:5116) > 2018-08-22 15:36:15,161+0200 INFO (libvirt/events) [virt.vm] > (vmId='b139a9b9-16bc-40ee-ba84-d1d59e5ce17a') CPU stopped: onIOError > (vm:6157) > 2018-08-22 15:36:15,162+0200 DEBUG (libvirt/events) > [virt.metadata.Descriptor] values: {'minGuaranteedMemoryMb': 1024, > 'clusterVersion': '4.2', 'resumeBehavior': 'au > to_resume', 'memGuaranteedSize': 1024, 'launchPaused': > 'false', 'startTime': 1534944832.058459, 'destroy_on_reboot': False, > 'pauseTime': 4999289.49} (metadata:596) > 2018-08-22 15:36:15,162+0200 DEBUG (libvirt/events) > [virt.metadata.Descriptor] values updated: {'minGuaranteedMemoryMb': 1024, > 'clusterVersion': '4.2', 'resumeBehavi > or': 'auto_resume', 'memGuaranteedSize': 1024, 'launchPaused': > 'false', 'startTime': 1534944832.058459, 'destroy_on_reboot': False, > 'pauseTime': 4999289.49} (metadat > a:601) > > 2018-08-22 15:36:15,168+0200 DEBUG (libvirt/events) > [virt.metadata.Descriptor] dumped metadata for > b139a9b9-16bc-40ee-ba84-d1d59e5ce17a: > <?xml version='1.0' encoding > metadata blablablabla............................> > > > > 2018-08-22 15:36:15,169+0200 DEBUG (libvirt/events) [virt.vm] > (vmId='b139a9b9-16bc-40ee-ba84-d1d59e5ce17a') event Suspended detail 2 > opaque None (vm:5520) > 2018-08-22 15:36:15,169+0200 INFO (libvirt/events) [virt.vm] > (vmId='b139a9b9-16bc-40ee-ba84-d1d59e5ce17a') CPU stopped: onSuspend > (vm:6157) > 2018-08-22 15:36:15,174+0200 WARN (libvirt/events) [virt.vm] > (vmId='b139a9b9-16bc-40ee-ba84-d1d59e5ce17a') device sda reported I/O > error (vm:4065) > 2018-08-22 15:36:15,340+0200 DEBUG (vmchannels) [virt.vm] > (vmId='46d496af-e2d0-4caa-9a13-10c624f265d8') Guest's message heartbeat: > {u'memory-stat': {u'swap_out': 0, > u'majflt': 0, u'swap_usage': 0, u'mem_cached': 119020, u'mem_free': > 3693900, u'mem_buffers': 2108, u'swap_in': 0, u'swap_total': 8257532, > u'pageflt': 141, u'mem_tota > l': 3878980, u'mem_unused': 3572772}, u'free-ram': u'3607', u'apiVersion': > 3} (guestagent:337) > > > Do you have some idea ? > _______________________________________________ > Users mailing list -- users@ovirt.org > To unsubscribe send an email to users-le...@ovirt.org > Privacy Statement: https://www.ovirt.org/site/privacy-policy/ > oVirt Code of Conduct: https://www.ovirt.org/community/about/community- > guidelines/ > List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/ > message/74HTXVBQEYTF4C5ZTH4FR3UQ3SM7LBQU/ >
_______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/QZIXQXM7NV7ELNQML4DGVNXPJEFWTDBZ/