collector: Add link_layer label to all per-port infiniband metrics#3666
Open
pallamidessi wants to merge 1 commit into
Open
collector: Add link_layer label to all per-port infiniband metrics#3666pallamidessi wants to merge 1 commit into
link_layer label to all per-port infiniband metrics#3666pallamidessi wants to merge 1 commit into
Conversation
…port metric exposed by the infiniband collector ## Context Modern Mellanox/NVIDIA NICs (ConnectX-6/7, BlueField, B200 platforms) routinely present some ports in Ethernet link layer (RoCE / generic Ethernet) while others run pure InfiniBand. Both modes appear under /sys/class/infiniband/* and currently emit indistinguishable node_infiniband_* series, so operators running mixed fleets cannot scope alerts (link_downed_total, link_error_recovery_total, etc.) easily `link_layer` is already parsed by [prometheus/procfs/sysfs into InfiniBandPort.LinkLayer](https://github.com/prometheus/procfs/blob/v0.20.1/sysfs/class_infiniband.go#L112) ## Testing Fixture i40iw0/ports/1/link_layer is set to "Ethernet" (iWARP runs over Ethernet by definition) so the e2e test exercises both link_layer values. Signed-off-by: Joseph Pallamidessi <joseph.pallamidessi@fluidstack.io>
pallamidessi
changed the title
link_layer label to all per-port metricslink_layer label to all per-port infiniband metrics
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds link_layer ("InfiniBand" or "Ethernet") as a label on every per port metric exposed by the infiniband collector
Context
Modern Mellanox/NVIDIA NICs (ConnectX-6/7, BlueField, B200 platforms) routinely present some ports in Ethernet link layer (RoCE / generic Ethernet) while others run pure InfiniBand. Both modes appear under /sys/class/infiniband/* and currently emit indistinguishable node_infiniband_* series, so operators running mixed fleets cannot scope alerts (link_downed_total, link_error_recovery_total, etc.) easily
link_layeris already parsed by procfs: prometheus/procfs#700Testing
Fixture i40iw0/ports/1/link_layer is set to "Ethernet" (iWARP runs over Ethernet by definition) so the e2e test exercises both link_layer values.
Alternative design
First time contributor here, as an alternative design, we can add a new
node_infiniband_port_info{device, port, link_layer}that could be used with a promQL join and is a more additive/contained change instead.@discordianfish