Thank you for the MySQL 5.4 Community Release
MySQL 5.4 was released this week at the MySQL conference, and billed as the “community release.” This seemed odd to some of us, and I heard lots of comment on it in the hallways. After all, the release was a surprise; the community didn’t know it was happening.
A source from within Sun, who is familiar with the details and wishes to remain anonymous, told me the story behind the release. I want to say that after this conversation, I fully support the release of 5.4 and I want to praise it generously. It is a great step in lots of right directions! It’s good for everyone. The secrecy was a necessary strategy.
The details must remain a secret, but perhaps someday it’ll be known.
Thank you Sun!
Can you at least tell us what kind of secret? :-P
Tim McCormack
27 Apr 09 at 10:50 pm
Thanks for the understanding Baron – we’re working hard to live up to the expectations of the community, and we’ve worked hard to push 5.4 to be a release that will meet them – I hope it does (though I’m pretty sure it will)!
Now, how we can start getting some of the work you guys are doing fed back in to it as well? :)
Cheers,
Mark
Mark Leith
28 Apr 09 at 5:40 am
Mark, honestly I think there’s a lot of other work to be done first.
Xaprb
28 Apr 09 at 9:00 am
Oh I agree of course, and we’re working hard to open up to the community as well (http://blogs.sun.com/datacharmer/entry/the_pursuit_of_openness).
What else would you guys need, to be able to start feeding back patches? There’s some great stuff in there, I think it’s a real shame that the “wider community” misses out, and would love to see you guys working closer with us now that we are trying to open up with such vigor.
Mark Leith
28 Apr 09 at 9:10 am
[...] is a little bit baffled, but Xapbr posts a Thank You for MySQL 5.4 Community. I think the baffling is due to several things: (1) There wasn’t much talk about MySQL 5.4 as [...]
Matt’s Blog » Blog Archive » Link Roundup
30 Apr 09 at 8:39 pm
We need the IO performance changes for InnoDB from the Percona and Google patches. 5.4 has support for more background IO threads, but a lot of work has been done in the last 6 months to make things much better.
It would be nice to know that performance stats aggregated by user and table are in the performance schema roadmap.
Mark Callaghan
30 Apr 09 at 9:00 pm
Hey Mark,
Yea, we know you want that pretty bad. :)
We currently have this worklog for looking at table/object usage:
http://forge.mysql.com/worklog/task.php?id=4895
Would this fit your needs for table monitoring?
Many of the core developers (Serg, Kostja, et al) are pretty keen to see the object monitoring rather than file IO as well, so there’s a lot of push to get this in there as well from inside too.
Vote on the worklog if it fits your needs! :)
Mark Leith
1 May 09 at 5:27 am
Have I read this correctly?
From http://forge.mysql.com/worklog/task.php?id=4895 they will generate events for every row operation and then from http://forge.mysql.com/worklog/task.php?id=4816 support will eventually be added for summary tables.
I want the summary tables first. I am skeptical that the per-event instrumentation can be done without a lot of CPU overhead. It will certainly have a lot of memory overhead to buffer events. And I want to have stats enabled all of the time for the summary tables.
Mark Callaghan
4 May 09 at 9:32 am