Koschei in production and staging have different Kerberos principals, so Koji staging sync script should update it.
I can't add attachment, so patch inline:
From 1e7dffb7113fcbc223c5cf23cd2c05a370992c91 Mon Sep 17 00:00:00 2001 From: Mikolaj Izdebski <mizdebsk@redhat.com> Date: Wed, 1 Mar 2017 10:48:50 +0100 Subject: [PATCH] Fix krb_principal for koschei user in staging Koji --- playbooks/manual/staging-sync/templates/koji-reset-staging.sql | 2 ++ 1 file changed, 2 insertions(+) diff --git a/playbooks/manual/staging-sync/templates/koji-reset-staging.sql b/playbooks/manual/staging-sync/templates/koji-reset-staging.sql index 260c47f..abbaaad 100644 --- a/playbooks/manual/staging-sync/templates/koji-reset-staging.sql +++ b/playbooks/manual/staging-sync/templates/koji-reset-staging.sql @@ -61,6 +61,8 @@ truncate host_channels; select now() as time, 'expiring repos' as msg; update repo set state = 3 where state in (0, 1, 2); +-- fix krb_principal for koschei user +update users set krb_principal='koschei/koschei-backend01.stg.phx2.fedoraproject.org' where name='koschei'; -- add our staging builders, dynamically pulled from ansible inventory -- 2.9.3
This has now been fixed in staging and the patch is merged for future syncs. Thank you!
Metadata Update from @puiterwijk: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Login to comment on this ticket.