Hello,

I've been a proudly user of mariadb 5.5.x for a long time now but given the nice feature setI decided to give mariadb 10.x a try.

I took one of our current slaves running mariadb 5.5.x (on centos 6.5) and followed the upgrade steps using yum and ran *mysql_upgrade*. The upgrade ran without any trouble...then I moved a huge table (about 1B records right now) from InnoDB to TokuDB.

Now, every couple of hours I find that the replication is far behind the master and the reason is because this server keeps checking tables marked as crashed....

I can't seems to find any indicator of OOM killer in the system logs NOR anything related to that in the mysql log, the only clue I have is:

---------------
150303 10:08:57 mysqld_safe Number of processes running now: 0

150303 10:08:57 mysqld_safe mysqld restarted

150303 10:08:58 [Warning] 'THREAD_CONCURRENCY' is deprecated and will be removed in a future release.

150303 10:08:59 [Warning] option 'innodb-status-file': boolean value 'ib_status' wasn't recognized. Set to OFF.

150303 10:08:59 [Warning] option 'innodb-autoextend-increment': unsigned value 10485760 adjusted to 1000

150303 10:08:59 [Note] InnoDB: Using mutexes to ref count buffer pool pages

150303 10:08:59 [Note] InnoDB: The InnoDB memory heap is disabled

150303 10:08:59 [Note] InnoDB: Mutexes and rw_locks use GCC atomic builtins

150303 10:08:59 [Note] InnoDB: Memory barrier is not used

150303 10:08:59 [Note] InnoDB: Compressed tables use zlib 1.2.3

150303 10:08:59 [Note] InnoDB: Using Linux native AIO

150303 10:08:59 [Note] InnoDB: Using CPU crc32 instructions

150303 10:08:59 [Note] InnoDB: Initializing buffer pool, size = 6.0G

150303 10:08:59 [Note] InnoDB: Completed initialization of buffer pool

150303 10:08:59 [Note] InnoDB: Highest supported file format is Barracuda.

150303 10:08:59 [Note] InnoDB: Log scan progressed past the checkpoint lsn 14178608217047

150303 10:08:59 [Note] InnoDB: Database was not shutdown normally!

150303 10:08:59 [Note] InnoDB: Starting crash recovery.

150303 10:08:59 [Note] InnoDB: Reading tablespace information from the .ibd files...

150303 10:09:05 [Note] InnoDB: Restoring possible half-written data pages

150303 10:09:05 [Note] InnoDB: from the doublewrite buffer...

InnoDB: Doing recovery: scanned up to log sequence number 14178613459456

InnoDB: Doing recovery: scanned up to log sequence number 14178618702336

InnoDB: Doing recovery: scanned up to log sequence number 14178623945216

InnoDB: Doing recovery: scanned up to log sequence number 14178629188096

InnoDB: Doing recovery: scanned up to log sequence number 14178634430976

InnoDB: Doing recovery: scanned up to log sequence number 14178639673856

InnoDB: Doing recovery: scanned up to log sequence number 14178644916736

......

150303 10:09:14 [Note] InnoDB: Starting an apply batch of log records to the database...

InnoDB: Progress in percent: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

InnoDB: Apply batch completed

InnoDB: In a MySQL replication slave the last master binlog file

InnoDB: position 0 16433109, file name slave-bin.852141

InnoDB: Last MySQL binlog file position 0 16784214, file name /var/lib/mysql/log/bin/slave-bin.005211

150303 10:09:34 [Note] InnoDB: 128 rollback segment(s) are active.

150303 10:09:34 [Note] InnoDB: Waiting for purge to start

150303 10:09:34 [Note] InnoDB: Percona XtraDB (http://www.percona.com) 5.6.22-72.0 started; log sequence number 14178834821658

Tue Mar 3 10:09:36 2015 TokuFT recovery starting in env /var/lib/mysql/data/

Tue Mar 3 10:09:36 2015 TokuFT recovery scanning backward from 1330297

Tue Mar 3 10:09:36 2015 TokuFT recovery bw_end_checkpoint at 1330297 timestamp 1425395277901416 xid 1330281 (bw_newer)

Tue Mar 3 10:09:36 2015 TokuFT recovery bw_begin_checkpoint at 1330281 timestamp 1425395273487081 (bw_between)

Tue Mar 3 10:09:36 2015 TokuFT recovery turning around at begin checkpoint 1330281 time 4414335

Tue Mar 3 10:09:36 2015 TokuFT recovery starts scanning forward to 1330297 from 1330281 left 16 (fw_between)

Tue Mar 3 10:09:36 2015 TokuFT recovery closing 14 dictionaries

Tue Mar 3 10:09:36 2015 TokuFT recovery making a checkpoint

Tue Mar 3 10:09:36 2015 TokuFT recovery done

150303 10:09:36 [Note] Recovering after a crash using /var/lib/mysql/log/bin/slave-bin

---------------

The odd thing is that the 5.5.x version was working just fine (since about a year) in the same hardware...nothing changed in that front.

Any clue?

Thank you in advance