Sunday, November 3, 2013

Teaser on my upcoming Percona Live London 2013 session

As you probably know already, I have a session on PERFORMANCE_SCHEMA at the conference, scheduled at 12 November 4:00pm - 4:50pm @ Orchard 1. Presentation is mostly ready, but I had not decided yet when to publish it. In the meantime, for those really interested, here is a teaser.

Below I list one link for each slide (in order of presentation) having more than one of them mentioned or listed in my notes. Now try to guess what I am going to say there and why. Note that it's not a tutorial (my half a day tutorial on PERFORMANCE_SCHEMA was not accepted by the conference committee, and this is probably good decisions, as I am usually very good in explaining what's bad or what should never be done and much worse in "best practices"). So, it's hardly a usual material on PERFORMANCE_SCHEMA...
  1. http://bugs.mysql.com/ - this is obvious, I am always speaking about bugs in any MySQL-related context.
  2. American Civil War - surely I plan to say something related to history
  3. Observer effect (physics) - ... and physic.
  4. How Percona does a MySQL Performance Audit - you always wanted to know this, aren't you?
  5. Science - yes, it's more science than art here, at Percona...
  6. http://www.juliandyke.com/Diagnostics/Events/EventReference.html - useful reference if you ever plan to move (back?) to good old Oracle RDBMS performance tuning
  7. SHOW INNODB STATUS walk through - I hope you've read this more than once. Maybe you had written something better on the topic? Then, please, share in the comment - I'd really want to read it one day.
  8. timed_mutexes system variable - do you know what is it for?
  9. http://dev.mysql.com/doc/refman/5.6/en/innodb-metrics-table.html - manual page for INFORMATION_SCHEMA.INNODB_METRICS table. So, I use references to the manual.
  10. http://dom.as/2013/04/17/on-performance-schemas/ - and I mention Domas and his posts more than once. That's good in any MySQL-related context.
  11. Bug #24795. This is one of many bugs mentioned. This time it's just a reminder about Jeremy Cole and what he did for MySQL many years ago.
  12. pt-pmp - as I work for Percona now, I surely have to mention Percona Toolkit (probably it's among my contract terms). This is not the only tool I mention.
  13. Bug #61545. Guess what this bug has to do with PERFORMANCE_SCHEMA in MySQL 5.6.
  14. Bug #69236. MySQL 5.6 is slow for single-threaded workloads. We all know this, so what?
  15. http://www.fromdual.ch/mysql-oprofile. Oli also has a session at PLUK 2013. It would be nice to meet him finally.
  16. http://dom.as/tag/dtrace/ - had you ever seen what Domas can do with DTrace on Mac OS X?
  17. http://marcalff.blogspot.com/ - not at the very beginning, but surely I mention a person who had actually implemented PERFORMANCE_SCHEMA...
  18. http://dev.mysql.com/doc/refman/5.6/en/performance-schema-instrument-naming.html - in some cases we just need a reference.
  19. http://mysqlentomologist.blogspot.com/2013/01/how-to-use-performanceschema-to-check.html - quoting myself, why not...
  20. Bug #68097. I've reported this bug, so I speak about it. Guess how many times.
  21. http://www.slideshare.net/Leithal/performance-schema-andpshelper - check this if you want to see some really good presentation on PERFORMANCE_SCHEMA. Mine has almost nothing in common with it.
  22. WL #2360. Do you know that some good old MySQL WorkLogs are still available in public?
  23. Why Performance Schema Overhead?.. - I have some ideas, but who can explain it better than DimitriK?
  24. Bug #69527. What, again a bug? This time it's a feature request and it's implemented in MySQL 5.7.3.
  25. Bug #68514. They say it's "Not a bug". Decide for yourself...
  26. Bug #68413. I wonder how many time I should mention this bug in public before it is closed somehow (not even fixed, just closed, somehow)...
  27. Bug #68855. It's a feature request, not directly related to PERFORMANCE_SCHEMA.
  28. Bug #68079. I've reported this bug and it surely shows how Oracle cares about MySQL.
 That's all. Now you know that there are least 28 slides in my presentation, that I am going to mention some names and speak a lot about bugs. If you want to check how all the above are related to real presentation, wait for the conference and attend my session (or read entire presentation soon). In any case it's now less than 10 days to wait.

If you want me to say (or NOT say) something special related to PERFORMANCE_SCHEMA, please, add a comment. I still have some time to add more slides or change my mind entirely...