Bug 1256021
| Summary: | [PPC64LE] Failed to add ppc host to engine | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Virtualization Manager | Reporter: | Artyom <alukiano> | ||||
| Component: | ovirt-engine | Assignee: | Michal Skrivanek <michal.skrivanek> | ||||
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Artyom <alukiano> | ||||
| Severity: | urgent | Docs Contact: | |||||
| Priority: | urgent | ||||||
| Version: | 3.6.0 | CC: | abologna, alukiano, gklein, hannsj_uhl, istein, lpeer, lsurette, mgoldboi, michal.skrivanek, mpoledni, rbalakri, Rhev-m-bugs, srevivo, ykaul | ||||
| Target Milestone: | ovirt-3.6.0-rc | Keywords: | AutomationBlocker, Regression, Triaged | ||||
| Target Release: | 3.6.0 | ||||||
| Hardware: | ppc64le | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | 3.6.0-12 | Doc Type: | Bug Fix | ||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2016-04-20 01:34:47 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | Virt | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Bug Depends On: | |||||||
| Bug Blocks: | 1201513, 1277183, 1277184 | ||||||
| Attachments: |
|
||||||
please attach engine.log... Created attachment 1066025 [details]
engine log
the problem is that the cpu model is reported as 'model_POWER8' and engine expects 'model_power8' to workaround this you can change the engine configuration to support this model (in the db) i verified with Artyom this works and we could run a vm. afaik, this was recently changed in libvirt, Michal, this a bug in libvirt or this is the correct name and we need to change the engine accordingly? This is intended change - libvirt reports what can be sent back to qemu's cpu element. 'power8' is not really a CPU model in qemu, on the other hand 'POWER8' is a magic constant that is used to select the best CPU choice for current version. We should treat POWER8 as the correct model for POWER8 and POWER8e (and possibly any other POWER machine at this point). Additional issue that I encounter with lower-uppercase, when we create vm we send cpu model with lowercase letters and libvirt just ignore cpu model because he not have such model.
Thread-38635::DEBUG::2015-09-02 03:15:50,888::__init__::534::jsonrpc.JsonRpcServer::(_serveRequest) Return 'VM.create' in bridge with {u'acpiEnable': u'true', u'emulatedMachine': u'pseries-rhel7.2.0', u'tabletEnable': u'true', 'pid': '0', u'memGuaranteedSize': 1024, u'transparentHugePages': u'true', 'displaySecurePort': '-1', u'spiceSslCipherSuite': u'DEFAULT', u'cpuType': u'power8',
VDSM currently doesn't create POWER8 cpu element at all so it isn't issue of upper/lower case. There is upstream patch posted to add this functionality and that one will expect 'POWER8' cpuType. Verified on rhevm-3.6.0-0.15.master.el6.noarch Success to add host to cluster with IBM POWER 8 cpu type |
Description of problem: Failed to add ppc host to engine with error: Host ***** moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all Version-Release number of selected component (if applicable): RHEVM: rhevm-3.6.0-0.12.master.el6.noarch HOST: vdsm-4.17.3-1.el7ev.noarch How reproducible: Always Steps to Reproduce: 1. Try to add host to engine 2. 3. Actual results: Failed with error above Expected results: Action success and host have state up Additional info: vdsClient -s 0 getVdsCaps clusterLevels = ['3.6'] cpuCores = '20' cpuFlags = 'powernv,model_POWER8' cpuModel = 'POWER8E (raw), altivec supported' cpuSockets = '4' cpuSpeed = '3690.000000' cpuThreads = '20' emulatedMachines = ['pseries', 'pseries-rhel7.2.0'] not sure about component problem or under backend or under vdsm