Weave seems to be failing some requests with 401s, consistently depending on username (we think). Perhaps this is related to a script Toby ran today to move people.
...move people to new nodes in preparation for the 0.5 server launch.
My 'thunder' user seems to be unaffected, but 'thundertest30' which I created earlier does not work.
anant is on the case
Assignee: server-ops → anant
We updated the primaryNodes for all users to be sj-weave01/06 this morning. The local ldap slave at sj-weave01 replicated only entries that had the attribute set to sj-weave01, and thus all other users who were "moved" (to nodes 2 through 6) started receiving 401 errors. I reverted all primaryNode attributes to sj-weave01, and it seems like the changes have been replicated to local LDAP server. All our test accounts work now.
Status: NEW → RESOLVED
Last Resolved: 9 years ago
Resolution: --- → FIXED
Slight update - we actually migrated all users to sj-weave2-6. Nobody was left on sj-weave01. I suspect the replication lag that caused us to miss this in the first place ultimately kept thunder from getting a login failure.
Created attachment 396527 [details] Weave Sync.log For me the issue still exist. Please look in the attachment.
Product: mozilla.org → mozilla.org Graveyard
You need to log in before you can comment on or make changes to this bug.