# upgrade 5.5 to 5.6 will not startup ERROR Aborting last message in error.log WHY? **URL:** \ **Category:** Other MySQL® Questions\ **Created:** [March 25, 2016, 1:27pm UTC](https://forums.percona.com/t/upgrade-5-5-to-5-6-will-not-startup-error-aborting-last-message-in-error-log-why/4807 "2016-03-25T13:27:59Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![MR.Y](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@MR.Y](https://forums.percona.com/u/MR.Y)\ **Post date:** [March 25, 2016, 1:27pm UTC](https://forums.percona.com/t/upgrade-5-5-to-5-6-will-not-startup-error-aborting-last-message-in-error-log-why/4807/1 "2016-03-25T13:27:59Z")
Why can I not startup up 5.6.29 on a CentOS 7 host where Percona 5.5.28 runs w/o problem Stop slave and shutdown 5.5.28 clean removed old symlink to mysql 5.5.28 binaries. NEXT yum install Percona-Server-shared-56-5.6.29-rel76.2.el6.x86\_64.rpm yum install Percona-Server-client-56-5.6.29-rel76.2.el6.x86\_64.rpm yum install Percona-Server-server-56-5.6.29-rel76.2.el6.x86\_64.rpm NEXT Want to startup Mysql first time and run mysql\_upgrade script # /usr/sbin/mysqld --skip-grant-tables --user=mysql & [1] 2051 ]# 2016-03-25 11:33:59 0 [Warning] TIMESTAMP with implicit DEFAULT value is deprecated. Please use --explicit\_defaults\_for\_timestamp server option (see documentation for more details). 2016-03-25 11:33:59 0 [Note] /usr/sbin/mysqld (mysqld 5.6.29-76.2-log) starting as process 2051 … 2016-03-25 11:33:59 2051 [ERROR] Can’t read from messagefile ‘/usr/share/mysql/english/errmsg.sys’ BUT it never starts up and there is not real info in the error.log only show ERROR Aborting … why not? 2016-03-25 11:12:30 1825 [Note] InnoDB: Waiting for purge to start 2016-03-25 11:12:30 1825 [Note] InnoDB: Percona XtraDB ([http://www.percona.com](http://www.percona.com)) 5.6.29-76.2 started; log sequence number 23270845141317 2016-03-25 11:12:30 1825 [ERROR] /usr/sbin/mysqld: unknown variable ‘table\_cache=4096’ 2016-03-25 11:12:30 1825 [ERROR] Aborting 2016-03-25 11:12:30 1825 [Note] Binlog end 2016-03-25 11:12:30 1825 [Note] Shutting down ---
**Author:** ![mirfan](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/mirfan/32/16890_2.png) [@mirfan](https://forums.percona.com/u/mirfan)\ **Post date:** [March 25, 2016, 4:00pm UTC](https://forums.percona.com/t/upgrade-5-5-to-5-6-will-not-startup-error-aborting-last-message-in-error-log-why/4807/2 "2016-03-25T16:00:32Z")
> > 2016-03-25 11:12:30 1825 [ERROR] /usr/sbin/mysqld: unknown variable 'table\_cache=4096 > > Renamed the table\_cache system variable to table\_open\_cache. Any scripts that refer to table\_cache should be updated to use the new name. So, replace table\_cache with table\_open\_cache in my.cnf and restart mysql server. It should fix the problem. I found this too [https://bugs.mysql.com/bug.php?id=68315](https://bugs.mysql.com/bug.php?id=68315) ---
**Author:** ![MR.Y](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@MR.Y](https://forums.percona.com/u/MR.Y)\ **Post date:** [March 25, 2016, 6:21pm UTC](https://forums.percona.com/t/upgrade-5-5-to-5-6-will-not-startup-error-aborting-last-message-in-error-log-why/4807/3 "2016-03-25T18:21:20Z")
MySQL was complaining on this 55 variables when trying to start 56 percona server So I comment them out - hence it should default to table\_open\_cache grep ‘##YE##’ /etc/my.cnf ##YE##character\_set\_client = utf8 ##YE##basedir = /usr/local/mysql ##YE##table\_cache = 4096 ##YE##innodb\_additional\_mem\_pool\_size = 256M The FIX was to change basedir from (55) basedir=/usr/local/mysql (Tar ball install and symlink install to /usr/local/mysql) to (56) basedir=/usr (rpm install of Percona server) then Mysql 5.6 server started up w/o problem and I could proceed with mysql\_upgrade script etc MySQL 56 slave is currently up running and attached to a 55 Master … mysql\> select @@table\_open\_cache; ±-------------------+ | @@table\_open\_cache | ±-------------------+ | 2000 | ±-------------------+ thx