There was some interesting discussion following my post last week asking whether there is a growing rift between commercial open source software vendors and some aspects of the open source user community.

Amongst the comments, Chris Marino of SnapLogic suggested that some of the tension might be eased by open source software vendors being more upfront about their intentions via the publication of social contracts. Examples include the Debian Social Contract and also Funambol’s Open Source Project Social Contract.

I know many of you may be wondering why Cool Stack 1.3 isn't out yet, so I thought I'd post an update on where we are with the release. When I first started talking about this, I was hoping to get the release out in early May. But several things have colluded to cause a delay.

1. We took on a lot for this release. Although just looking at the proposal, it may seem like a simple matter to just update the versions of the various components, underneath the hood, we decided to do a lot more streamlining of the build and packaging process.  Since so many version updates were involved, that also meant more legal approvals.
2. We have a more formal QA process for this release (and for future releases). The good news is that this means we will have a more quality release, but the bad news is that it now takes more time as we have to …

I know many of you may be wondering why Cool Stack 1.3 isn't out yet, so I thought I'd post an update on where we are with the release. When I first started talking about this, I was hoping to get the release out in early May. But several things have colluded to cause a delay.

1. We took on a lot for this release. Although just looking at the proposal, it may seem like a simple matter to just update the versions of the various components, underneath the hood, we decided to do a lot more streamlining of the build and packaging process.  Since so many version updates were involved, that also meant more legal approvals.
2. We have a more formal QA process for this release (and for future releases). The good news is that this means we will have a more quality release, but the bad news is that it now takes more time as we have to …

We (in the Telecom team at MySQL) have been debating whether we should call this blog 'MySQL in Telco' or 'MySQL in Communications'. Naming discussions tend to take long time, and this one was no exception.

From a US perspective, it appears that Telecom is an outdated term. Wireless carriers and cable television companies do not consider themselves as Telecom companies. Maybe for this reason, large US-headquartered vendors including Sun, HP, IBM and Oracle all have a 'Communications & Media' practice.

From a European perspective, Telecom is used for equipment vendors and service providers. Companies like Logica, Cap Gemini, Atos Origin and TietoEnator refer to the vertical as 'Telecom & Media'.

Of the global SI's in India, Wipro and TCS refer to Telecom while Infosys talk about Communications.

As a working title we at one time used 'MySQL blablabla' blog, and funnily enough, there was a compromise suggestion to …

Quite frequently I would log in to customers system and find MySQL using too much memory. I would look at memory consumed by Innodb (it is often higher than innodb_buffer_pool_size) substract memory used by other global buffers such as query_cache_size and key_buffer and will in many cases see some mysterous memory which I can't really explain. It can be several Gigabytes accounting for over 50% of memory usage of MySQL in some cases, though typically it is much smaller fraction.

Now. There are a lot of guesses I can make. Could it be memory allocated for per connection buffers which was not really freed to operation system ? Could it be some buffers associated with opened tables, prepared statements etc ? Could it be some of the queries currently running is using some of the buffers temporary based on per connection settings ? It also could be memory leak in MySQL or runaway stored …

MySQL 6.0 Feature #2: Online Backup

Online Backup
MySQL 6.0 New Features Document#2
2008-05-18

By Alexander Nozdrin, Chuck Bell, Lars Thalmann, Peter Gulutzan, Rafal Somla

High Executive Summary

Online backup is a new feature in MySQL 6.0.
It protects from database loss.
It does not block all other concurrent MySQL connections.
It will interest Database Administrators.

Database Administrators Summary

BACKUP DATABASE copies all data and metadata in one or more MySQL databases, into an “image file”. RESTORE reads an image file and rewrites all the data and metadata in one or more MySQL databases. So if you lose a database, you can recover all of it as of the time of the last BACKUP DATABASE statement.And then you can re-run the statements in MySQL’s binary log to recover “from the time of …

I can find most of MySQL’s “timeout” variables with a SHOW statement:

mysql> show variables like '%timeout%';+----------------------------+-------+

| Variable_name              | Value |

+----------------------------+-------+

| connect_timeout            | 10    |

| delayed_insert_timeout     | 300   |

| falcon_lock_wait_timeout   | 50    |

| innodb_lock_wait_timeout   | 50    |

| ...................................|

| interactive_timeout        | 28800 |

| net_read_timeout           | 30    |

| net_write_timeout          | 60    |

| slave_net_timeout          | 3600  |

| table_lock_wait_timeout    | 50    |

| wait_timeout               | 28800 |

+----------------------------+-------+

11 rows in set (0.01 sec)

The straightforward list conceals several anomalies and buried controversies.

* Should we allow non-integer values like 0.55 seconds, or perhaps make all values milliseconds rather than seconds, in order to …

In the earlier article about upcoming features, “Roadmap”, there was no mention of materialized views. The explanation is simple: they’re not on the roadmap. Yet.

However, this is one of the things that might change. There is a strong desire for materialized views, and one of MySQL’s architects has a strong interest (he wrote his PhD thesis on the topic). So if their absence from the “Roadmap” article troubled you, despair not. Keep watching this space and maybe there will be further news someday.

For those who haven’t memorized the names of every symbol: the tilde is the wavy line that appears occasionally over specific characters, particularly over N in Spanish. But it can appear alone. If you’re North American and you look at the top left corner of your keyboard, you’ll probably see it:

~

Anyway, we ran into this bug during a comprehensive re-test of all the Chinese character sets:

mysql> create table tbig5 (s1 char(5) character set big5);

Query OK, 0 rows affected (0.11 sec)mysql> insert into tbig5 values ('Y'),('~');

Query OK, 2 rows affected (0.08 sec)

Records: 2  Duplicates: 0  Warnings: 0

mysql> select * from tbig5 where s1 = 'Y';

+------+

| s1   |

+------+

| Y    |

| ~    |

+------+

2 rows in set (0.04 sec)

This is Bug#25420 “Tilde = ‘Y’ in Chinese”. It has been around for years, ever since the contribution of the BIG5 Chinese character set. No Chinese …

The verdict is in: DTrace will be on by default.

This is a wildly different thing from saying “DTrace will be there”. No right-minded person would want to eschew the benefits of DTrace, the performance-analysis tool par excellence from Sun Microsystems for all varieties of Solaris operating systems. Even if you aren’t a Solaris user, I’d suggest you have a quick glance at the literature about it because it’s the concept that non-Sun folk will surely try to imitate someday. And who knows, maybe you’ll become a Solaris user just because of this. I’ve heard of a consultant in California who does Oracle gigs thus: he goes into the shop where they’re running (say) some other Unix variant, asks them to copy their data to a Solaris machine that he handily brings with him, and re-runs their problem queries with DTrace to find the bottlenecks.

But the controversy (if I can call it that) was over whether the standard MySQL …