# Problems while upgrading Percona cluster from 5.6 to 5.7 due to innodb\_log\_block\_size

**URL:** <https://forums.percona.com/t/problems-while-upgrading-percona-cluster-from-5-6-to-5-7-due-to-innodb-log-block-size/6383>\
**Category:** Other MySQL® Questions\
**Created:** [May 26, 2018, 2:57am UTC](https://forums.percona.com/t/problems-while-upgrading-percona-cluster-from-5-6-to-5-7-due-to-innodb-log-block-size/6383 "2018-05-26T02:57:15Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ethaniel](https://avatars.discourse-cdn.com/v4/letter/e/ba8739/32.png) [@ethaniel](https://forums.percona.com/u/ethaniel)\
**Post date:** [May 26, 2018, 2:57am UTC](https://forums.percona.com/t/problems-while-upgrading-percona-cluster-from-5-6-to-5-7-due-to-innodb-log-block-size/6383/1 "2018-05-26T02:57:15Z")

</div>

Hello everyone!

I have 3 servers running Percona 5.6 which are in a cluster.

I have innodb\_log\_block\_size=4096 set on each one of them. The disks are SSD running ext4 filesystems formatted at 512 bytes. I chose innodb\_log\_block\_size=4096 because several manuals mentioned that it’s a better option for SSD disks.

I tried to perform a in-place upgrade from 5.6 to 5.7 (according to [https://www.percona.com/doc/percona-…ide\_56\_57.html](https://www.percona.com/doc/percona-server/LATEST/upgrading_guide_56_57.html)).

It failed because innodb\_log\_block\_size is set to 4096 (innodb\_log\_block\_size was removed in 5.7):

> [@](#):
>
> [ERROR] InnoDB: Upgrade after a crash is not supported. This redo log was created before MySQL 5.7.9, and we did not find a valid checkpoint. Please follow the instructions at [http://dev.mysql.com/doc/refman](http://dev.mysql.com/doc/refman)  
> /5.7/en/upgrading.html

The manual ([https://www.percona.com/doc/percona-…ged\_in\_57.html](https://www.percona.com/doc/percona-server/LATEST/changed_in_57.html)) states that I need to shutdown mysql, remove the variable from my.cnf configuration, remove the old logs and start the service again before doing the upgrade.

But since I’m running in a cluster, won’t it cause problems during the upgrade? If I shutdown mysql, remove this option and remove the logs, won’t the logs be syncronised back from other nodes once I start mysql back up?