Hi,
I've recently updated to latest VDSM code (commit c506391442bc996031aa871a4dfd61368b5a30db). Ever since, whenever I run a VM, my host moves to Non-Operational state (oVirt).
* getVdsStats output [1]: doesn't include any network sessions.
* Error from oVirt engine log: - "Host '120' moved to Non-Operational state because interface/s 'em1,' are down which needed by network/s 'ovirtmgmt, ' in the current cluster"
* Reverting latest commit (http://gerrit.ovirt.org/#/c/13838) seems to solve the issue.
Thanks, Daniel
[1] vdsClient 0 getVdsStats: cpuIdle = '100.00' cpuSys = '0.00' cpuSysVdsmd = '0.00' cpuUser = '0.00' cpuUserVdsmd = '0.00' dateTime = '2013-04-14T08:27:01 GMT' elapsedTime = '1465' generationID = 'c6c1a126-48d6-44a1-8bd7-80648436a615' ksmCpu = 0 ksmPages = 100 ksmState = False memAvailable = 11064 memCommitted = 577 memShared = 0 momStatus = 'active' netConfigDirty = 'False' rxRate = '0.00' statsAge = '1463.61' storageDomains = {'064a6bef-39f4-4bb4-a811-b8206fa7dd1c': {'code': 0, 'delay': '0.0126550197601', 'lastCheck': '3.9', 'valid': True}, '14a320f9-06c8-41b9-bb53-4fc3dc5717b0': {'code': 0, 'delay': '0.0133948326111', 'lastCheck': '3.9', 'valid': True}, 'd2066c79-a6f9-41e8-8ac8-642f84e415c0': {'code': 0, 'delay': '0.0102300643921', 'lastCheck': '7.2', 'valid': True}} swapFree = 7952 swapTotal = 7983 txRate = '0.00' vmActive = 1 vmCount = 1 vmMigrating = 0
On 04/14/2013 08:11 PM, Daniel Erez wrote:
Hi,
I've recently updated to latest VDSM code (commit c506391442bc996031aa871a4dfd61368b5a30db). Ever since, whenever I run a VM, my host moves to Non-Operational state (oVirt).
getVdsStats output [1]: doesn't include any network sessions.
Error from oVirt engine log: - "Host '120' moved to Non-Operational state because interface/s 'em1,' are down which needed by network/s 'ovirtmgmt, ' in the current cluster"
Reverting latest commit (http://gerrit.ovirt.org/#/c/13838) seems to solve the issue.
Thanks, Daniel
Daniel,
Thanks for reporting the problem. It's my fault! That commit is not self-contained in the branch because of code merge mistake. I appologize for it.
@Dan, I think your quick fix also resolves the problem of net statistics missing. The break causes StatsThread.run() exit with no sample records stored, and then the get() return fake stats for other fields in the case of length of samples is less than 2. But we don't that trick for each interface's statistics. It's also verified from my test.
Mark.
[1] vdsClient 0 getVdsStats: cpuIdle = '100.00' cpuSys = '0.00' cpuSysVdsmd = '0.00' cpuUser = '0.00' cpuUserVdsmd = '0.00' dateTime = '2013-04-14T08:27:01 GMT' elapsedTime = '1465' generationID = 'c6c1a126-48d6-44a1-8bd7-80648436a615' ksmCpu = 0 ksmPages = 100 ksmState = False memAvailable = 11064 memCommitted = 577 memShared = 0 momStatus = 'active' netConfigDirty = 'False' rxRate = '0.00' statsAge = '1463.61' storageDomains = {'064a6bef-39f4-4bb4-a811-b8206fa7dd1c': {'code': 0, 'delay': '0.0126550197601', 'lastCheck': '3.9', 'valid': True}, '14a320f9-06c8-41b9-bb53-4fc3dc5717b0': {'code': 0, 'delay': '0.0133948326111', 'lastCheck': '3.9', 'valid': True}, 'd2066c79-a6f9-41e8-8ac8-642f84e415c0': {'code': 0, 'delay': '0.0102300643921', 'lastCheck': '7.2', 'valid': True}} swapFree = 7952 swapTotal = 7983 txRate = '0.00' vmActive = 1 vmCount = 1 vmMigrating = 0
vdsm-devel@lists.fedorahosted.org