collector: add nvmesubsystem collector for NVMe-oF path health#3579
Open
sradco wants to merge 2 commits intoprometheus:masterfrom
Open
collector: add nvmesubsystem collector for NVMe-oF path health#3579sradco wants to merge 2 commits intoprometheus:masterfrom
sradco wants to merge 2 commits intoprometheus:masterfrom
Conversation
sradco
force-pushed
the
add_collector_multipath
branch
from
742c1b1 to
a0a146e
Compare
Author
|
Hi @SuperQ , I created this PR for a new multipath collector. |
sradco
force-pushed
the
add_collector_multipath
branch
from
a0a146e to
1fa2099
Compare
sradco
changed the title
sradco
force-pushed
the
add_collector_multipath
branch
from
1fa2099 to
635b613
Compare
sradco
changed the title
jsafrane
reviewed
Mar 13, 2026
There was a problem hiding this comment.
The code looks good to me. I tested it with 2 NVMe over TCP devices, I got:
node_nvmesubsystem_info{iopolicy="numa",model="Linux",nqn="tempdisk",serial="bab529a0f32e397e1319",subsystem="nvme-subsys0"} 1
node_nvmesubsystem_path_state{controller="nvme0",state="connecting",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme0",state="dead",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme0",state="live",subsystem="nvme-subsys0",transport="tcp"} 1
node_nvmesubsystem_path_state{controller="nvme0",state="resetting",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme0",state="unknown",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme1",state="connecting",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme1",state="dead",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme1",state="live",subsystem="nvme-subsys0",transport="tcp"} 1
node_nvmesubsystem_path_state{controller="nvme1",state="resetting",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_path_state{controller="nvme1",state="unknown",subsystem="nvme-subsys0",transport="tcp"} 0
node_nvmesubsystem_paths_live{subsystem="nvme-subsys0"} 2
node_nvmesubsystem_paths_total{subsystem="nvme-subsys0"} 2
Which looks reasonable.
Comment on lines
+64
to
+70
| switch raw { | ||
| case "live", "connecting", "resetting", "dead": | ||
| return raw | ||
| case "deleting", "deleting (no IO)", "new": | ||
| return raw | ||
| default: | ||
| return "unknown" |
There was a problem hiding this comment.
for the record, I checked that this is a complete list of all states reported by the kernel today.
sradco
force-pushed
the
add_collector_multipath
branch
3 times, most recently
from
51fb644 to
8acec17
Compare
sradco
force-pushed
the
add_collector_multipath
branch
2 times, most recently
from
17875ea to
d7d2b27
Compare
sradco
force-pushed
the
add_collector_multipath
branch
from
d7d2b27 to
1a4cac6
Compare
Author
|
PR is failing since its based on prometheus/procfs#797, which is still not merged |
sysfs metrics Add a new disabled-by-default collector that reads /sys/block/dm-* to discover Device Mapper multipath devices and expose path health metrics. Multipath devices are identified by checking that dm/uuid starts with "mpath-", which distinguishes them from LVM or other DM device types. The path state is reported as-is from /sys/block/<dev>/device/state, supporting both SCSI devices (running, offline, blocked, etc.) and NVMe devices (live, connecting, dead, etc.) without hardcoding a fixed set of states. All device-level metrics include both the DM friendly name (device) and the kernel block device name (sysfs_name, e.g. dm-0) to enable direct correlation with node_disk_* I/O metrics without recording rules. No special permissions are required — the collector reads only world-readable sysfs attributes. Exposed metrics: - node_dmmultipath_device_info - node_dmmultipath_device_active - node_dmmultipath_device_size_bytes - node_dmmultipath_device_paths - node_dmmultipath_device_paths_active - node_dmmultipath_device_paths_failed - node_dmmultipath_path_state Signed-off-by: Shirly Radco <sradco@redhat.com> Co-authored-by: AI Assistant <noreply@cursor.com>
Add a new disabled-by-default collector that reads /sys/class/nvme-subsystem/ to expose NVMe over Fabrics subsystem connectivity metrics. This complements the existing nvme collector (which reports per-controller hardware stats) by monitoring the subsystem-level path redundancy — how many controller paths are live, connecting, or dead for each NVMe subsystem. Exposed metrics: - node_nvmesubsystem_info - node_nvmesubsystem_paths - node_nvmesubsystem_paths_live - node_nvmesubsystem_path_state Signed-off-by: Shirly Radco <sradco@redhat.com> Co-authored-by: AI Assistant <noreply@cursor.com>
sradco
force-pushed
the
add_collector_multipath
branch
from
04fd251 to
705c86e
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add a new disabled-by-default collector that reads
/sys/class/nvme-subsystem/ to expose NVMe over Fabrics subsystem
connectivity metrics.
This complements the existing nvme collector (which reports
per-controller hardware stats) by monitoring the subsystem-level
path redundancy - how many controller paths are live, connecting,
or dead for each NVMe subsystem.
Exposed metrics:
Depends on prometheus/procfs#797
Signed-off-by: Shirly Radco sradco@redhat.com
Co-authored-by: AI Assistant noreply@cursor.com