Back to bug 2172808

Who When What Removed Added
Red Hat One Jira (issues.redhat.com) 2023-02-23 07:21:44 UTC Link ID Red Hat Issue Tracker RHCEPH-6174
Xiubo Li 2023-02-23 07:23:24 UTC Target Release 6.1z1 5.3z2
Xiubo Li 2023-02-23 07:23:57 UTC Assignee vshankar xiubli
Vivek Das 2023-02-27 05:07:52 UTC CC vdas
Severity unspecified medium
Xiubo Li 2023-03-29 04:37:04 UTC Status NEW ASSIGNED
Link ID Ceph Project Bug Tracker 58853 Ceph Project Bug Tracker 58825
Venky Shankar 2023-03-29 12:25:32 UTC CC vshankar
Status ASSIGNED POST
errata-xmlrpc 2023-03-31 14:20:25 UTC CC tserlin
Flags needinfo?(xiubli)
Flags needinfo?(vshankar)
Flags needinfo?(xiubli)
Flags needinfo?(vshankar)
CC vereddy
Flags needinfo?(vereddy)
Status POST MODIFIED
Status MODIFIED ON_QA
Ken Dreyer (Red Hat) 2023-03-31 14:20:59 UTC Flags needinfo?(vereddy)
CC kdreyer
Amarnath 2023-04-03 12:22:37 UTC CC amk
Flags needinfo?(xiubli)
Amarnath 2023-04-05 19:06:51 UTC Status ON_QA VERIFIED
Xiubo Li 2023-04-10 01:21:57 UTC Flags needinfo?(xiubli)
Akash Raj 2023-04-10 14:30:34 UTC CC akraj
Docs Contact akraj
Flags needinfo?(xiubli)
Xiubo Li 2023-04-11 00:42:46 UTC Flags needinfo?(xiubli)
Doc Type If docs needed, set a value Enhancement
Doc Text Feature:

Convert the client requests forwarding and retrying counters to _u32 type from __u8 type. And will limit the forwarding and retrying to 256 times.

Reason:

In multiple active MDSs case if one request fails in curruent MDS for some reason the client will forward the request to another MDS, while actually if no MDS could successfully handle it the request will bounce infinitly between MDSs. The old num_fwd/num_retry counters are __u8 type, which will overflow after bouncing 256 times.

Result:

The client requests will stop forwarding and retrying after certain 256 times and will fail directly instead of infinitely forwarding and retrying.
Akash Raj 2023-04-11 05:37:32 UTC Blocks 2185621
Akash Raj 2023-04-11 07:36:43 UTC Doc Text Feature:

Convert the client requests forwarding and retrying counters to _u32 type from __u8 type. And will limit the forwarding and retrying to 256 times.

Reason:

In multiple active MDSs case if one request fails in curruent MDS for some reason the client will forward the request to another MDS, while actually if no MDS could successfully handle it the request will bounce infinitly between MDSs. The old num_fwd/num_retry counters are __u8 type, which will overflow after bouncing 256 times.

Result:

The client requests will stop forwarding and retrying after certain 256 times and will fail directly instead of infinitely forwarding and retrying.
.Client request counters are converted from `_u8` type to `_u32` type and the limit is set to 256 times

Previously, in multiple active MDSs cases, if a single request failed in the current MDS, the client would forward the request to another MDS. If no MDS could successfully handle the request, it would bounce infinitely between MDSs. The old `num_fwd`/`num_retry` counters are `_u8` type, which would overflow after bouncing 256 times.

With this enhancement, the counters are converted from `_u8` type to `_u32` type and the limit for forwarding and retrying is set to 256 times. The client requests stop forwarding and retrying after 256 times and fails directly instead of infinitely forwarding and retrying.
errata-xmlrpc 2023-04-11 16:06:38 UTC Status VERIFIED RELEASE_PENDING
errata-xmlrpc 2023-04-11 20:07:59 UTC Status RELEASE_PENDING CLOSED
Resolution --- ERRATA
Last Closed 2023-04-11 20:07:59 UTC
errata-xmlrpc 2023-04-11 20:08:56 UTC Link ID Red Hat Product Errata RHBA-2023:1732
Ranjini M N 2023-05-19 07:10:58 UTC Blocks 2203283
Blocks 2203283

Back to bug 2172808