Bug #65660
closedmds: drop client metrics during recovery
0%
Description
When the rank is coming up, there's little reason to record historical metrics from the clients. We've also seen floods of these metrics messages slow down up:rejoin significantly.
Updated by Xiubo Li almost 2 years ago
Is this new in the upstream master ? As I remembered we have improved this and the clients will only send the metrics when the MDS is ready, which is in active state.
Updated by Christopher Hoffman almost 2 years ago
there's little reason to record historical metrics from the clients
Can you expand on this? Are we losing anything by dropping them?
Updated by Patrick Donnelly almost 2 years ago
Xiubo Li wrote in #note-1:
Is this new in the upstream master ? As I remembered we have improved this and the clients will only send the metrics when the MDS is ready, which is in active state.
That may be but the MDS should be resilient to older clients.
Updated by Patrick Donnelly almost 2 years ago
Christopher Hoffman wrote in #note-2:
there's little reason to record historical metrics from the clients
Can you expand on this? Are we losing anything by dropping them?
The metrics are there to provide a real-time view of performance for clients. Jos or Venky may correct me if I'm wrong but I don't think there is any use for past metrics.
Updated by Venky Shankar almost 2 years ago
Patrick Donnelly wrote in #note-4:
Christopher Hoffman wrote in #note-2:
there's little reason to record historical metrics from the clients
Can you expand on this? Are we losing anything by dropping them?
The metrics are there to provide a real-time view of performance for clients. Jos or Venky may correct me if I'm wrong but I don't think there is any use for past metrics.
Unless the metrics are persisted by the exporter daemon, which is currently in the works. So, we do not loose anything by dropping them.
Updated by Dhairya Parmar almost 2 years ago
Venky Shankar wrote in #note-5:
Patrick Donnelly wrote in #note-4:
Christopher Hoffman wrote in #note-2:
there's little reason to record historical metrics from the clients
Can you expand on this? Are we losing anything by dropping them?
The metrics are there to provide a real-time view of performance for clients. Jos or Venky may correct me if I'm wrong but I don't think there is any use for past metrics.
Unless the metrics are persisted by the exporter daemon, which is currently in the works. So, we do not loose anything by dropping them.
So you mean It'd be fine to have past metrics once they are being persisted by the exporter daemon? So once the exporter daemon is operational, we'd need to have this code adjusted again, right?
Updated by Patrick Donnelly over 1 year ago
- Category set to Correctness/Safety
- Status changed from In Progress to Pending Backport
- Backport changed from squid,reef to squid,reef,quincy
Updated by Upkeep Bot over 1 year ago
- Copied to Backport #66620: squid: mds: drop client metrics during recovery added
Updated by Upkeep Bot over 1 year ago
- Copied to Backport #66621: quincy: mds: drop client metrics during recovery added
Updated by Upkeep Bot over 1 year ago
- Copied to Backport #66622: reef: mds: drop client metrics during recovery added
Updated by Upkeep Bot over 1 year ago
- Tags (freeform) set to backport_processed
Updated by Upkeep Bot 8 months ago
- Merge Commit set to 426beef37ec414b84001d601ca9cb2febb628f1f
- Fixed In set to v19.3.0-2917-g426beef37ec
- Upkeep Timestamp set to 2025-07-09T13:46:26+00:00
Updated by Upkeep Bot 8 months ago
- Fixed In changed from v19.3.0-2917-g426beef37ec to v19.3.0-2917-g426beef37e
- Upkeep Timestamp changed from 2025-07-09T13:46:26+00:00 to 2025-07-14T17:11:31+00:00
Updated by Upkeep Bot 5 months ago
- Released In set to v20.2.0~2644
- Upkeep Timestamp changed from 2025-07-14T17:11:31+00:00 to 2025-11-01T00:57:42+00:00
Updated by Upkeep Bot 4 months ago
- Status changed from Pending Backport to Resolved
- Upkeep Timestamp changed from 2025-11-01T00:57:42+00:00 to 2025-11-20T13:09:19+00:00