# Changing skip\_log\_bin on a v5.7 Percona Cluster

**URL:** <https://forums.percona.com/t/changing-skip-log-bin-on-a-v5-7-percona-cluster/21779>\
**Category:** Percona XtraDB Cluster 5.x\
**Created:** [May 2, 2023, 3:25am UTC](https://forums.percona.com/t/changing-skip-log-bin-on-a-v5-7-percona-cluster/21779 "2023-05-02T03:25:03Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![jimbo881](https://avatars.discourse-cdn.com/v4/letter/j/9d8465/32.png) [@jimbo881](https://forums.percona.com/u/jimbo881)\
**Post date:** [May 2, 2023, 3:25am UTC](https://forums.percona.com/t/changing-skip-log-bin-on-a-v5-7-percona-cluster/21779/1 "2023-05-02T03:25:03Z")

</div>

I have a vendor supplied Percona 3 node cluster v5.7. They have skip\_log\_bin set so binlogs are not captured. I want to turn binlogs on for all 3 nodes but my team and I do not have a whole lot of experience with Percona Clusters. I am thinking I remove this variable from the my.cnf file on all 3 systems, then shut down node 3, 2 then 1…then fire back up 1, 2 then 3. Will this work without having to bootstrap node 1 again? Will they re-connect?

I will add expire\_logs\_days (yes I know it is deprecated) on all 3 nodes plus expire them once XtraBackup runs…

Any help would be greatly appreciated!

---

<div class="post-metadata">

**Author:** ![kedarpercona](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/kedarpercona/32/809_2.png) [@kedarpercona](https://forums.percona.com/u/kedarpercona)\
**Post date:** [May 2, 2023, 4:32am UTC](https://forums.percona.com/t/changing-skip-log-bin-on-a-v5-7-percona-cluster/21779/2 "2023-05-02T04:32:34Z")

</div>

Welcome back @jimbo881,

So we’re talking about Percona XtraDB Cluster 5.7. Find the steps below I’d try making this change.

- edit my.cnf

```auto
  server-id=1
  log-bin=mysql1-bin
  binlog_format=row
  log-slave-updates = ON
  expire_logs_days = 7

```

(note: adjust server-id and log-bin name as per your server)

- restart mysql
- Once the node joins back to cluster proceed further with next cluster node.  
`show global status like 'wsrep%';`

Thanks,  
Kedar