Date: 2008-04-11 09:46 Subject: Bug#27628 revisited. Deliver code, not bureaucracy! Security: Public

I reported "Processlist shows status NULL when server is purging/managing query cache" in April 2007, while I was still in MySQL support engineering. This issue actually made tracking down other bugs and customer issues take longer. About a year later, MySQL 6.0 codebase apparently has the progress info. By the way, we are essentially talking about a one-line fix, adding something like thd->proc_info="Flushing table from cache"; at the appropriate spot.

In the mean time, three bugfix meetings have spent time on this item (judging by the tags), those are meetings where MySQL (re-)prioritises some bugs for fixing. That's an awful lot of time to waste euh spend on this thing. The bug entry itself is still open as "feature request", and I presume MySQL support and bug engineers are still wasting time due to the insufficient reporting in 5.0 and 5.1. In the latest comment one developer suggests closing it (since it's in 6.0). But...

Why not just put the one-liner into 5.0 and 5.1 and be done with it? There can't possibly be any adverse functional impact to this fix. Well not that that's hindered a meriad of other 5.0 "fixes" from creeping into the world, but this one should really be harmless! And it would actually help users and support engineers. Deliver code already, today!

9 Comments | Post A Comment | Add to Memories | Tell a Friend | Link

Date: 2008-04-11 10:19 Subject: Hierarchies in SQL? OQGRAPH engine backend preview Security: Public

Dealing with hierarchies in a relational database is a pest. There's Oracle's CONNECT BY PRIOR, and the SQL standard defines IBM's recursive UNION, but still... wouldn't it be nice if a hierarchy (or even a full-on graph, like social networks have) could just be managed cleanly relationally?

This is something Kim and I have been dabbling with. The engine is called OQGRAPH (OQ for Open Query) now because just graph caused some symbol conflict hassles. Anyway, following is a brief demo of how it works.

First, let's insert some data...

mysql> INSERT INTO gstest (origid,destid) VALUES (1,2),(1,3),(2,4);

Get it back out plain...

mysql> SELECT origid AS node, destid AS edge FROM gstest;
+-------------+
| node | edge |
+-------------+
|    1 |    3 |
|    1 |    2 |
|    2 |    4 |
|    3 | NULL |
|    4 | NULL |
+-------------+

(Hey, notice something? the engine automatically added nodes 3 and 4: because there's edges going there, they must exist!)

Now for a more fancy trick: show me any (not shortest per-se, just any) path from 1 to 4 (2 hops)...

mysql> SELECT GROUP_CONCAT(linkid ORDER BY seq) AS path FROM gstest
    ->  WHERE latch=1 AND origid=1 AND destid=4;
+-------+
| path  |
+-------+
| 1,2,4 |
+-------+

Same from 3 to 2...

mysql> SELECT GROUP_CONCAT(linkid ORDER BY seq) AS path FROM gstest
    ->  WHERE latch=1 AND origid=3 AND destid=2;
Empty set

(Presuming uni-directional edges here, there is no possible path from 3 to 2)

How does it work? What else can it do? Does it make my morning coffee? Can I get it?
Well, for more info, background and discussion, come to the "dealing with hierarchies and graphs" BoF at the MySQL Conf, or contact me directly.

Oh, and you may ask: what's with the "any path" thing? Well, that was easier to implement than a shortest path algorithm, being 2am at the time ;-) But the system could do shortest path, and whatever else you might need.

Post A Comment | Add to Memories | Tell a Friend | Link

Date: 2008-04-11 14:25 Subject: MySQL Conf 2008 special: auction of unique signed T-shirt Security: Public

Goodness, I thought I'd never find it. Moving a few times causes a lot of havoc. But, here it is:

So, what is it, who's on it, and what are my intentions with it?

It's a 2002 summer (Northern hemisphere) edition MySQL t-shirt ("Power & Freedom" theme), size S. It was signed at the MySQL Cancun dev staff meeting, October 2003. This was the first dev meeting with the NDB Cluster team, and assorted others. I got this shirt signed by all present (except I notice I didn't actually sign it myself ;-)
Let's see if I can decode the names...

David Axmark (co-founder), Mikael Ronstrom (NDB architect), Pekka Nousiainen (NDB dev), Lenz Grimmer (build engineer, now community relations), Mike Zinner (GUI dev lead), Joao Prado Maia (Eventum), Jorge del Conde (dev,support), Kaj Arno (VP of lots), Megan Wagner (event management, Matt's wife), Greg Weir (Novell dev), Carsten Pederson (certification manager), a name I can't make out right now, Brian Aker (server architecture), Mark Matthews (Connector/J), Lars Thalman (NDB docs, now replication dev), Tobias Asplund (trainer/consultant), Harrison Fisk (trainer, now support engineering), Magnus Svensson (NDB engineer), Paul DuBois (docs magician), Stefan Hinz (docs lead), Matt Wagner (dev manager), Tom Basil (support director), and Monty himself.

The two drawings are by My and Max, Monty's oldest kids and the alleged source of the product naming. And that, my friends, makes this t-shirt rather special! You want to own this one-of-a-kind piece of history, don't you? Now's your chance: it will be auctioned during the MySQL Quiz night at the MySQL Conference 2008 (Wednesday 7-8.30pm).

What happens with the proceeds? I think EFF is a worthy target that most people will approve of. And for that reason I won't let it go cheap. Once you own it, it's up to you to decide if you want any others (including myself) to sign it, for now I will just keep it safe and bring it to the conf.

Post A Comment | Add to Memories | Tell a Friend | Link