mysql process (or node) restart triggers full Galera sync

[Brad Durrow]

{"build_id": "2013-12-27_00-24-14", "ostf_sha": "83ada35fec2664089e07fdc0d34861ae2a4d948a", "build_number": "214", "nailgun_sha": "af1598bcc9faf468d4d9265cc5c51fa8cea53136", "fuelmain_sha": "17eed776b30886851ae0042fa7a30184f5cd8eb6", "astute_sha": "6ce36837882399e0d3bb1ffdb2c3b2d8dcb84b54", "release": "4.0", "fuellib_sha": "eebe07913ee09311c8e7c9231f6785081327dc0e"}

Fuel 4.0, HA (3 controllers), Neutron VLAN, Ceph backend

A controller reboot or crm resource restart p-mysql causes that node to go into this state (as reported by mysql -e "show status like 'wsrep_local_state_comment';"):

wsrep_local_state_comment | Joining: receiving State Transfer

Currently on my cluster the node remains in this state for about 30 minutes.

[Vladimir Kuklin (vkuklin)]

Please, attach diagnostic snapshot.

[Andrew Woodward (xarses)]

Vlad, what logs would be helpful, Brad's environment has over 10gb of logs so a snapshot is not realistic.

[Vladimir Kuklin (vkuklin)]

mysql and all pacemaker logs are mandatory

[Mike Scherbakov (mihgen)]

Brad, thank you for reporting a bug. Unfortunately, looks like we can't proceed without logs. If you reproduce this again, please attach logs and any additional details you may fine useful. Closing as Invalid for now.

[Andrew Woodward (xarses)]

This is the expected process for Galera, however I'm marking it as a dupe of 1293680 because it being used in the VIP while this state is a problem

[Brad Durrow (l-brad)]

This is two separate problems. 1293673 compounds 1293680 but one is related to MySQL galera config and the other is related to haproxy config.

Please excuse my thumb's tpyos.

> On May 7, 2014, at 6:07 PM, Andrew Woodward <email address hidden> wrote:
>
> *** This bug is a duplicate of bug 1293680 ***
> https://bugs.launchpad.net/bugs/1293680
>
> This is the expected process for Galera, however I'm marking it as a
> dupe of 1293680 because it being used in the VIP while this state is a
> problem
>
> ** This bug has been marked a duplicate of bug 1293680
> while Galera node is in Sync or Donor state many services are down
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1293673
>
> Title:
> mysql process (or node) restart triggers full Galera sync
>
> Status in Fuel: OpenStack installer that works:
> Invalid
>
> Bug description:
> {"build_id": "2013-12-27_00-24-14", "ostf_sha":
> "83ada35fec2664089e07fdc0d34861ae2a4d948a", "build_number": "214",
> "nailgun_sha": "af1598bcc9faf468d4d9265cc5c51fa8cea53136",
> "fuelmain_sha": "17eed776b30886851ae0042fa7a30184f5cd8eb6",
> "astute_sha": "6ce36837882399e0d3bb1ffdb2c3b2d8dcb84b54", "release":
> "4.0", "fuellib_sha": "eebe07913ee09311c8e7c9231f6785081327dc0e"}
>
> Fuel 4.0, HA (3 controllers), Neutron VLAN, Ceph backend
>
> A controller reboot or crm resource restart p-mysql causes that node
> to go into this state (as reported by mysql -e "show status like
> 'wsrep_local_state_comment';"):
>
> wsrep_local_state_comment | Joining: receiving State Transfer
>
> Currently on my cluster the node remains in this state for about 30
> minutes.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/fuel/+bug/1293673/+subscriptions