I am running percona xtradb cluster 8.0. One of my nodes went down, when I try to bring it up. The log says the transfer is complete and joined.

0.0 (DB404): State transfer from 3.0 (DB403) complete.
[Galera] Shifting JOINER -> JOINED (TO: 799642)

However looking at the wsrep_local_state_comment status says joining. The wsrep_cluster_size has 5 which is the total amount including the bad node. Also showing all incoming ip addresses as well. wsrep_ready is off but I am guessing that is because it is still joining.
The log file also fills up with

[InnoDB] Page cleaner took 11974ms to flush 474 pages

And the node never fully joins the cluster. When I try to query I get WSREP has not yet prepared node for application use Any advice would be appreciated.

Page cleaner took ms to flush pages/WSREP has not yet prepared node for application use

I just let it run for about 22 hours and it finally has fully joined. Page clearer still fires every so often. No real resolution, it just started working. I am thinking it some sort of hardware related issue, or could be bad software install.

  • What kind of app? What kind of schema? (eg EAV) How many queries per second?
    – Rick James
    Commented Sep 15, 2022 at 20:27
  • I am just trying to join the cluster. No one can query the node. Commented Sep 15, 2022 at 22:36
  • What is going on in the other nodes?
    – Rick James
    Commented Sep 16, 2022 at 0:51
  • The other nodes are working as expected. No issues at all. I checked just now and page cleaner message is still running and the bad node still has joining as its status. The wsrep_local_recv_queue is going down. It was at 900 yesterday now it is at 112. Commented Sep 16, 2022 at 11:38


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.