mirror of
https://github.com/prometheus-community/windows_exporter.git
synced 2026-10-17 18:19:04 +02:00
Jan-Otto KröpkeandClaude
9015955cfa
Signed-off-by: Claude <noreply@anthropic.com> Co-authored-by: Claude <noreply@anthropic.com>
53 lines
4.2 KiB
Markdown
53 lines
4.2 KiB
Markdown
# smbclient collector
|
|
The smbclient collector collects metrics from MS SmbClient hosts through perflib
|
|
|||
|
|
-|-
|
|
Metric name prefix | `smbclient`
|
|
Data source | Performance Data
|
|
Counters | `SMB Client Shares`
|
|
Enabled by default? | No
|
|
|
|
## Flags
|
|
|
|
None
|
|
|
|
## Metrics
|
|
Name | Description | Type | Labels
|
|
-----|-------------|------|-------
|
|
`windows_smbclient_data_queue_seconds_total` | Seconds requests waited on queue on this share | counter | `server`, `share`|
|
|
`windows_smbclient_read_queue_seconds_total` | Seconds read requests waited on queue on this share | counter | `server`, `share`|
|
|
`windows_smbclient_write_queue_seconds_total` | Seconds write requests waited on queue on this share | counter | `server`, `share`|
|
|
`windows_smbclient_request_seconds_total` | Seconds waiting for requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_stalls_total` | The number of requests delayed based on insufficient credits on this share | counter | `server`, `share`|
|
|
`windows_smbclient_requests_queued` | The point in time (current) number of requests outstanding on this share | gauge | `server`, `share`|
|
|
`windows_smbclient_data_bytes_total` | The bytes read or written on this share | counter | `server`, `share`|
|
|
`windows_smbclient_requests_total` | The requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_metadata_requests_total` | The metadata requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_read_bytes_via_smbdirect_total` | The bytes read from this share via RDMA direct placement | counter | `server`, `share`|
|
|
`windows_smbclient_read_bytes_total` | The bytes read on this share | counter | `server`, `share`|
|
|
`windows_smbclient_read_requests_via_smbdirect_total` | The read requests on this share via RDMA direct placement | counter | `server`, `share`|
|
|
`windows_smbclient_read_requests_total` | The read requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_turbo_io_reads_total` | The read requests that go through Turbo I/O | counter | `server`, `share`|
|
|
`windows_smbclient_turbo_io_writes_total` | The write requests that go through Turbo I/O | counter | `server`, `share`|
|
|
`windows_smbclient_write_bytes_via_smbdirect_total` | The written bytes to this share via RDMA direct placement | counter | `server`, `share`|
|
|
`windows_smbclient_write_bytes_total` | The bytes written on this share | counter | `server`, `share`|
|
|
`windows_smbclient_write_requests_via_smbdirect_total` | The write requests to this share via RDMA direct placement | counter | `server`, `share`|
|
|
`windows_smbclient_write_requests_total` | The write requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_read_seconds_total` | Seconds waiting for read requests on this share | counter | `server`, `share`|
|
|
`windows_smbclient_write_seconds_total` | Seconds waiting for write requests on this share | counter | `server`, `share`|
|
|
|
|
Windows keeps several raw counters at 32 bits: the time sums behind `read_seconds_total`, `write_seconds_total` and `request_seconds_total`, and the counts behind `requests_total`, `read_requests_total`, `write_requests_total`, `metadata_requests_total`, `stalls_total`, `turbo_io_reads_total`, `turbo_io_writes_total`, `read_requests_via_smbdirect_total` and `write_requests_via_smbdirect_total`. The collector extends them past 2^32 per share, so they don't drop back to a low value when the raw counter wraps. The extended value starts at the raw value and follows the raw counter when the share reconnects.
|
|
|
|
The time sums wrap after 2^32 ticks, about 430 seconds of summed request time, and that sum grows by one second per second for each outstanding request. The collector can only detect one wrap between two scrapes, so with about 7 or more requests outstanding on average over a 60 second scrape interval, the `*_seconds_total` values undercount. Scrape more often on busy shares.
|
|
|
|
## Useful queries
|
|
```
|
|
# Average request queue length (includes read and write).
|
|
irate(windows_smbclient_data_queue_seconds_total)
|
|
# Request latency milliseconds (includes read and write).
|
|
irate(windows_smbclient_request_seconds_total) / irate(windows_smbclient_requests_total) * 1000
|
|
```
|
|
## Alerting examples
|
|
_This collector does not yet have alerting examples, we would appreciate your help adding them!_
|
|
|