[ZBXNEXT-6847] Add CSVFS volume discovery Created: 2021 Aug 23 Updated: 2024 Aug 15 |
|
| Status: | Confirmed |
| Project: | ZABBIX FEATURE REQUESTS |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Change Request | Priority: | Trivial |
| Reporter: | Antons Sincovs | Assignee: | Zabbix Development Team |
| Resolution: | Unresolved | Votes: | 3 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Description |
|
CSVFS (Cluster Shared Volumes File System) volumes are created and configured within Microsoft Windows Server failover cluster or Azure Stack HCI and mounted on nodes by the path: C:\ClusterStorage\Volume<IncrementalNumber> Such volumes cannot be correctly discovered for later use with performance counters, as for referring to the corresponding performance metrics the disk drive number is needed, however the filesystem discovery rule: vfs.fs.discovery returns only the full path and the file system type, but not the disk drive number (as seen in the example):
vfs.fs.discovery [s|[{"{#FSNAME}":"C:","{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed"},{"{#FSNAME}":"D:","{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed"},{"{#FSNAME}":"E:","{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed"},{"{#FSNAME}":"F:","{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed"},{"{#FSNAME}":"C:\\ClusterStorage\Volume1","{#FSTYPE}":"CSVFS","{#FSDRIVETYPE}":"fixed"}]]
The disk drive number can be obtained using the performance instance discovery rule: perf_instance.discovery[physicaldisk] which returns the drive number, but in its turn lacks the filesystem type:
perf_instance.discovery[physicaldisk] [s|[{"{#INSTANCE}":"0 C: D:"},{"{#INSTANCE}":"1 Q:"},{"{#INSTANCE}":"2"},{"{#INSTANCE}":"_Total"}]]
In order to obtain the required performance metric we need to refer to it the following way: perf_counter_en[\PhysicalDisk(2)\Avg. Disk sec/Read] Thus it is needed to get both filesystem type and disk drive number within one discovery rule. Here is the required structure of the discovery rule as seen by the client (which is essentially vfs.fs.disovery appended with {#INSTANCE} taken from perf_instance.discovery):
vfs.fs.discovery
[s|[
{"
{#FSNAME}":"C:","{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed","{#INSTANCE}":"0 C: D:"},
{"{#FSNAME}
":"D:","
{#FSTYPE}":"NTFS","{#FSDRIVETYPE}":"fixed","{#INSTANCE}":"0 C: D:"},
{"{#FSNAME}":"E:","{#FSTYPE}
":"NTFS","
{#FSDRIVETYPE}":"fixed","{#INSTANCE}":"2"},
{"{#FSNAME}":"C:\\ClusterStorage
Volume1","{#FSTYPE}":"CSVFS","{#FSDRIVETYPE}
":"fixed","
{#INSTANCE}
":"3"},
]]
Suggested community solutions declined by the client:
|
| Comments |
| Comment by Punster [ 2023 Jan 18 ] |
|
Support for this would be a game changer for us. Multiple Windows cluster with CSV |