Francesco Romani has posted comments on this change.
Change subject: caps: Collect numa information ......................................................................
Patch Set 5: Code-Review+1
(2 comments)
http://gerrit.ovirt.org/#/c/23703/5//COMMIT_MSG Commit Message:
Line 15: The items are separated by semicolon, the fields in each item Line 16: are separated by colon. Line 17: Take the first item above to explain the meaning of each field: Line 18: 0 - The index of numa node Line 19: 0,1 - The index of cpus(core) which belong to this numa node ok, so is cpu,core right? Line 20: 10240 - The total memory of this numa node Line 21: 9500 - The free memory of this numa node Line 22: Line 23: Change-Id: I63eeb697ab986c3b9cad0dc44f41924f329e52cd
http://gerrit.ovirt.org/#/c/23703/5/vdsm/caps.py File vdsm/caps.py:
Line 181: memInfo = _getMemoryStatsByNumaCell(int(cellInfo[0]), 0, memStats) Line 182: cellInfo.append(str(memInfo['total']/1024)) Line 183: cellInfo.append(str(memInfo['free']/1024)) Line 184: cellsInfo.append(':'.join(cellInfo)) Line 185: return ';'.join(cellsInfo) I feel this can be improved somehow but I don't have any suggestion here. On the other hand, we also have XML processing improvements in the pipeline and I don't see any obvious issue, in the nd I think it is good enough to go. Line 186: Line 187: Line 188: def _getMemoryStatsByNumaCell(cell, flags=0, memStats=None): Line 189: if memStats is None: