Everything is a Freaking DNS problem

When I read Monty's post on leaving this passage struck me the most.

The main reason for leaving was that I am not satisfied with the way the MySQL server has been developed, as can be seen on my previous blog post. In particular I would have like to see the server development to be moved to a true open development environment that would encourage outside participation and without any need of differentiation on the source code. Sun has been considering opening up the server development, but the pace has been too slow.

In short, Sun isn't open enough. I think I've said that enough, it's typically more Open Core than Open Source .. and for a growing amount of people.. that isn't good enough.

Reacting on that post we see Matt Asay trying to convince his CNet Audience that Open Source Ideals don't translate well into a big sofware business.

I think Tarus view I see it as just the opposite. Open source spells the end of big software, if big software is defined as companies that make billions of dollars from selling software licenses. is much more to the point.

And the way Sun has been working, with MySQL, Virtualbox and different others doesn't seem to work that well either.
I stopped counting the P companies but I think I`m allmost a point where I know mere ex-Sun/MySQL employees than current Sun/MySQL employees and it's not just on the MySQL side that this happens.

Maybe Monty leaving Sun will wake up the powers that be, maybe it won't.

Finishing off with another quote from Tarus

In my mind Monty is a role model and I wish him all the best.

Larry, I`m glad you asked ..

With the risk of receiving a flood of comments pointing me to already existing tools here's my go at what I as a sysadmin of often large deployments am looking for in Drupalland .

You suggest LDAP and syslog integration .. guess we already have that don't we ?

But what I haven't found yet ..
An rpm/deb repository of Drupal modules So we can do an apt-get install drupal-package , yes I know about Drush, but I want the files on my system to be in a package and clearly identified, it helps me keeping my system uncluttered.

apt-get update drupal-package , or yum install drupal-package would be a luxury, same for themes btw.

It would be lovely if the postinstalls of those package also trigger a database upgrade if needed.

Which brings me to the next issue. If I have a multisite setup and I update a module in sites/all/modules, I usually have to go trough each and every site hosted there to update the databases. No really something I like to do for 10+ euh 3+ sites.

Also think big, don't waste your time on desktop apps or guis .since as long as you only have 1 site to manage point an click is fine, for you , Think in terms of what if you have 10 sites, 100 sites, or more .. do you really want to do that kind of administration via a browser or gui ? Some wise man once said If your computer can't install it the installer is broken. A script should or automation tool should be able to interact with the sites, not a human operating a mouse ;) It's not just the RSI , but also the fact that to err is human, and if the computer fails a script you can patch it :)

If you read this blog, you should probably sign the petition !

There is a really easy way to figure out ...

Look at the size of of the devrooms , if at all , a project gets at Fosdem
It's really interresting to see the Embedded room move to a 500 seat room which it really needed when the first openMoko talk was held there in a previous edition.

The FDO , Drupal, Mozilla, Centos/Fedora and Suse rooms stay in similar size rooms as last year.. But it seems lik the BSD and PostgreSQL room which was pretty crowded moved to a bigger location.

The Ruby room also seems to move to a bigger room. Fosdem has a couple of new rooms too it seems this freed up room for new groups such as the MySQL Crowd

Obviously these sizes aren't a real match to the size of a community, as the new rooms might need (and probably will need) a reshuffle for next year ;)

Different schedules for the devrooms are online .. you'll probably find me in the Debian , Fedora+CentOS, MySQL or Drupal rooms :)

Did I mention that :

and I`ll actually be speaking there again about Monitoring MySQL

A while ago Dries wondered about Contributing back to Drupal .

Now Inuits is not a Webshop, we are an Open Source shop, so you won't see a zillion Drupal modules being contributed by us in the near future (albeit there are a couple) , we are company assisting other organisations in their adoption of Open Source, and Drupal is amongst the projects we care for. You'll notice code from us in the different other Open Source projects, including the Linux Kernel and other core infrastructure. And not all of it was code, there is a lot of published documentation, methodologies, bughunting and also spreading the words, or talking about our experiences around different topics such as MySQL Cluster and Drupal, or different alternatives to Monitor , or to Monitor MySQL etc ..

But we've been contributing in different other ways to open source and we have always been benefiting from that. And we try to convince our customers to do so too.
So Dries is right in all his reasons why one should contribute back to open source, specially as an organisation that uses OpenSource for it's customers you just have to.

The lack of contributions however might have different reasons
I can imagine however that the moment a commercial Open Source company starts backing or distributing people start looking at that project in a different way.

The amount of contributions to a pure open source project has been traditionally lower than the amount of code contributed to an open core project.

And I`m pretty sure that most of the braindead box moving RedHat and other so called Value Added Resellers aren't contributing a single line of code and my fear is that with the current growth of open source adoption that more of those traditional IT shops will just resell support subscriptions as if it were just another software product.

Now the Drupal community is one with a lot of developers so things might be different there compared to a group of system integrators that are installing operating systems and don't know the difference between python and php.

I think we'll see more and more open source users that aren't planning to contribute back, (although it is easy ) , But do we honestly think te group of developers is infinite ? However the more Open Source users the more chance we have these users turn into developers, I just don't think the current percentages will stay the same.

Anyway .. we'll continue to contribute.. and we hope you do the same ...

My MythTV store lives on an LVM volume that is spread over 2 disks, one of them is an external USB disk. So the cleaninglady seems to have touched a cable and after coming back from holiday I had a read-only filesystem that afer a remount had about 350Gb in lost+found with irrelevant filenames.

total 337407844
drwx------   2 tv   tv         4096 Dec 17 22:47 .
drwxrwxrwx  15 tv   tv         4096 Dec 17 22:44 ..
-rw-r--r--   1 root root  423343556 Dec 14 07:10 I303109.RCN
-rw-r--r--   1 root root 2990538924 Dec 13 19:05 I303107.RCN
-rw-r--r--   1 root root 1023691768 Dec 13 08:10 I319494.RCN
-rw-r--r--   1 root root 1023622348 Dec 13 07:45 I327684.RCN
-rw-r--r--   1 root root  423735892 Dec 13 07:10 I327682.RCN
-rw-r--r--   1 root root  466749476 Dec 12 15:43 I135169.RCN
-rw-r--r--   1 root root 1023314212 Dec 12 07:45 I098309.RCN
-rw-r--r--   1 root root 1022459928 Dec 12 06:35 I098306.RCN
-rw-r--r--   1 root root 2458822948 Dec  9 22:50 I139264.RCN
-rw-r--r--   1 root root 2129683736 Dec  9 21:30 I323592.RCN
-rw-r--r--   1 root root  466735992 Dec  9 15:43 I323590.RCN
-rw-r--r--   1 root root 1022747296 Dec  9 07:45 I323588.RCN

Obviously I wanted to recover my data.
So I had files with a wrong filename on a filesystem but with a correct timestamp and probably the right filesize.
Luckily the mythconverg.recorded table also gives me lots of information about the files that mythtv had originally stored the content in.

mysql> select basename,lastmodified,filesize from recorded limit 10;
+---------------------------+---------------------+------------+
| basename                  | lastmodified        | filesize   |
+---------------------------+---------------------+------------+
| 1003_20081003230000.mpg   | 2008-10-07 21:53:49 | 6197765380 |
| 1093_20080320232600.mpg   | 2008-03-20 23:25:31 |          0 |
| 1075_20060301191300.mpg   | 2006-03-24 22:48:42 |          0 |
| 1002_20080729160500.mpg   | 2008-07-29 19:20:30 | 3679223940 |
| 592251_20081217072000.mpg | 2008-12-17 07:20:02 |          0 |
| 1002_20080911143500.mpg   | 2008-09-11 16:41:44 | 3486101572 |
| 1002_20080923143500.mpg   | 2008-09-23 16:49:41 | 3679789684 |
| 1033_20081110153500.mpg   | 2008-11-10 15:47:12 |  338877000 |
| 1002_20080922144000.mpg   | 2008-09-22 16:47:38 | 3485505140 |
| 1002_20080721160500.mpg   | 2008-07-23 20:52:16 | 3679559444 |
+---------------------------+---------------------+------------+
10 rows in set (0.00 sec)

My first idea was to use the mysql filesystem engine to query the filesytem and write me a simple query however I totally failed to build that engine :(
(Anyone else successfull here ? )

So I created a temp table

CREATE TABLE `temp2` (
  `size` bigint(20) default NULL,
  `oldname` varchar(255) default NULL,
  `lastmod` timestamp NOT NULL default CURRENT_TIMESTAMP on update CURRENT_TIMESTAMP
) ENGINE=MyISAM DEFAULT CHARSET=latin1

And parsed the content of my lost+found directory into a set of insert statements

ls -l --time-style=long-iso | awk  -F' ' '{print "insert into temp2 values ("  $5 ",\""  $8"\",\""$6" "$7"\");"}'

From there is a matter of grabbing the matching filenames

echo "select \"mv \" , oldname, basename from recorded, temp2 where temp2.size= recorded.filesize ;" | mysql mythconverg

And moving the actual files ... now all is back to normal ..

So I wanted a fresh MySQL 5.1 release on my Fedora 10 box.
Fedora comes with a mysql 5.0 package and the MySQL community packages for 5.1 conflict with that in different ways. (to start with MySQL vs mysql)

For some weird reason libmysqlclient (in mysql-libs) is required by redland, which is required by openoffice.org-core. I probaly never need the RDF funtionalities that redland provides me but I`m stuck with them anyhow.

So Fred pointed me to

Remi repo for mysql 5.1

yum --enablerepo remi install  mysql-server

Now gives me a very nice MySQL 5.1 to play around with.

The earlier posted fix , getting my data from mythportal seemed to be a temporary solution.

It also seems the solution that was mentioned at Kristof Vanhertum's blog , the agent fix doesn't work anymore either.

Somewhere in the comments of that post however people start referring to mc2xml a tool to get your info (even 14 days worth of it) from a really untrustworthy source and import it amongst other tools into MythTV

No source code to be found, just a binary .. something that always scares me .. But running the thing (first time as a dummy empty user, one never knows) actually seems to give me an XML file with relevant data , to be used by mythfilldatabase.

 ./mc2xml  -c be -g 2610 
Loading ..... : mc2xml <mc2xml@gmail.com> (2008-09-15)
Connecting .. : microsoft.com
 
Select lineup: 
 
  0: TER  Unmapped Analogue Terrestrial
  1: TER  Unmapped Digital Terrestrial
  2: CAB  Unmapped Analogue Cable
  3: CAB  Unmapped Digital Cable
  4: SAT  Unmapped Analogue Satellite
  5: SAT  Unmapped Digital Satellite
  6: SAT  TV Vlaanderen Digital Satellite
  7: TER  TV Vlaanderen Digital terrestrial
  8: CAB  Telenet Analog Kabel (132)
 
(0-8): 8 
 
Downloading . : microsoft.com
Downloaded .. : 353944 bytes
Source ...... : BDS
Headend ..... : 7200116
Version ..... : 2693590756
Start time .. : 2008-10-23T00:00:00 UTC
End time .... : 2008-11-06T23:55:00 UTC
Parsing ..... : Telenet Analog Kabel (132)
Writing ..... : xmltv.xml
Writing ..... : mc2xml.dat
Completed ... : 5384193 bytes in 26s, 27 channels, 9409 programs scheduled.

Yes that's my actual postcode. Different postcode values will give you other alternatives such as Coditel/Integan .. so you better use the right one.

I`m not really satisfied however, it fails to recognize my already defined channels. So I seem to have something that at first sight looks like relevant data . Only my channel numbers are way way way off ..

I need to fix that asap. The only problem being that tomorrow it's Geekdinner and I`ll be spending the rest of the weekend at T-Dose hence I have no time to fix this anymore ...

Anyone got a quick fix for that ? Or will I just have to live with a weekend of non-recordings ? :)

Well.. maybe next week will be a productive one then :)

To save Solaris from a certain death ?

Reading Planet MySQL the last couple of hours I'm trying really hard to convince myselve the Solaris offensive there is not orchestrated.. but I can't.
It might ofcourse be the fresh MySQL users that Sun brought in on their platform that started out blogging but hey .. I`m paranoia right :)

Are they really trying to get at least a fraction of the MySQL community on Solaris. Do they really think they can ? Yes they lost a zillion of Solaris customers that were running a proprietary database to MySQL on Linux users ,, but why would they want to move back to a semi proprietary setup ?

According to Linuxjournal Alan Cox seems to think that ZFS is the only thing that is keeping Solaris alive. I don't think DTrace was a bigg mass tool that would convince the crowds to suddenly move to an other operating system.

So is Sun trying to Lock In a community ? Other people would call it a Jail, altough I wouldn't consider this particular type of jail a Luxury one :)

Why can't Sun just drop the whole idea of building its own OS and contribute more , they are slowly learning in some other fields , but in the Operating System field they still haven't realised they should up the fight ... but then again .. I overheard someone say recently "The best fights to watch, are the ones that can't be won anymore .."

So just remember .. the L in LAMP still stands for Linux,

Matt Asay is pushing his favorite Open Source model again. The model where the majority of developers of a project work for a company and that company is creating a business around the project. There's nothing wrong with that model, but he seems to forget the other models time over time :)

Matt is absolutely right with 2 of the 3 things he wants you to consider.
A SI in the middle of a $50 million dollar project involving Alfresco not talking to Alfresco is just wrong. An SI not offering a support contract is also just wrong. But an SI forcing his customer to buy the commercially supported version from a vendor ? Where's the customer choice ?

The customer should have the option to choose for a commercially supported version or the free version. And preferably that should be an educated option.

Matt seems to forget about situations where an Open Source project is not managed by one central organization , by one company that contributes most of the code. What companies are in charge of Apache, Linux (and don't reply RedHat here), Xen (No it's not Citrix anymore) , Samba , and lots of others.

If you were in Australia why wouldn't you get a MySQL support contract from Arjen Lenz ? Even if he didn't have MySQL Partner Certification ?

If you were in Germany , as a Centos or even RHEL user would you want to get your critical Samba Support from some support guy at RedHat or from some German guy at a local shop

If Michael Badger were in the SI business , would he be a good partner to support your Zenoss setup ?

3 totally different cases, the ex-employee, the developer not employed by a vendor, the guy who wrote the book.

Not all Open Source projects are backed by 1 clearly identifyable company, lots of open source developers work at SI's and they might be a better source for a specific project than a vendor that just integrated their product.

Worse even .. I've seen tons of traditional SI's jump on the Open Source wagon, by working only with the Commercially backed Open Source tools, as if they were proprietary software. Obviously the commercial Open Source vendors love these SI's they are the best resellers , and probably the worst integrators.

So Matt, please remember, there is more open source on this planet than your corporate backed open source, I haven't seen figures , but my bet would be that the corporate backed part is the smallest one.

In the end the most important thing is that the customer has got to have the educated choice between the locally supported opensource alternative , or the locally supported opensource alternative with commercial backing ..

But then again , it might be the European vs US vision however :)