Comments on: XtraDB feature: save / restore buffer pool

By: serbaut

serbaut — Mon, 20 Sep 2010 12:28:17 +0000

@Vadim: yes you can.

By: Vadim

Vadim — Wed, 08 Sep 2010 14:48:18 +0000

serbaut,

Great, thanks!
Can we use it under BSD license ?

By: serbaut

serbaut — Wed, 08 Sep 2010 13:13:43 +0000

Patch against 5.1.47 for sorting before restore: http://gist.github.com/570107

By: Vadim

Vadim — Wed, 24 Mar 2010 23:21:13 +0000

Vojtech,

Thank you,
I posted
https://bugs.launchpad.net/percona-xtradb/+bug/546496

By: Vojtech

Vojtech — Sat, 06 Mar 2010 20:35:02 +0000

Vadim, I just recognized, I cannot kill the XTRA_LRU_RESTORE query. I think it should check for ‘killed status’ at least each few seconds. Am I right?

By: Patrick Mulvany

Patrick Mulvany — Thu, 28 Jan 2010 15:56:53 +0000

Vadim, How is this patch effected by increases or decreases in InnoDB buffer pool size parameter?
I would assume that changing this would not be a good idea.

By: Vadim

Vadim — Wed, 27 Jan 2010 16:38:53 +0000

Tobias,

Basically yes.
You need though to sort dump of buffer_pool, so it will be loaded sequentially.
Also dumping back to RAM may take some time (4 mins in my experiment), but it faster then work with cold cache.

By: Tobias Petry

Tobias Petry — Wed, 27 Jan 2010 12:43:26 +0000

Vadim, this means we can make a BufferPool-Dump, run the server 2 hours, crash it and after restarting (and innodb’s recovery) and dumping the BufferPool back to the RAM we have an warm InnoDB instance with no stale data? And no other problems?

By: Vadim

Vadim — Mon, 25 Jan 2010 18:53:05 +0000

Baron,

There is no reason why it will not work.

I as said we store just pointers to pages (space_id, page_id),
and at restore stage we just read pages by pointers.
It does not matter if InnoDB crashed before, did recovery procedure, etc.

By: Baron Schwartz

Baron Schwartz — Mon, 25 Jan 2010 04:18:55 +0000

Vadim, how does this interact with recovery? Does it work OK if you save the buffer pool contents, then crash the server, restart it, and restore the buffer pool? If not, then that might be a problem for the DRBD use case.

I think it should work fine, but maybe I’m wrong.