General update: Search troubles and solutions

It appears that the world is slowly discovering MusicBrainz. Last month we pushed nearly 2M pages per day and smoke is pouring out of our poor search server. Fortunately we have some money in the bank so, the board of directors approved the purchase a new RAM-heavy server for use as our new search server. … Continue reading “General update: Search troubles and solutions”

It appears that the world is slowly discovering MusicBrainz. Last month we pushed nearly 2M pages per day and smoke is pouring out of our poor search server. Fortunately we have some money in the bank so, the board of directors approved the purchase a new RAM-heavy server for use as our new search server. Then yesterday, Matthew Mullenweg of WordPress fame donated $5000 to the MetaBrainz Foundation!! That nicely covers the costs of our new search server (~$3000) and then some.

I’m also working on improving the search server so that it runs more efficiently and that we get fewer timeout errors. Things improved yesterday when I redirected all web-based searches to our new server. I hope to cut over the rest of the search traffic (web service traffic) to a better suited server later this afternoon.

In other words — I’m working hard to make sure the site stays responsive, even under the growing loads. Things should improve much more once we can deploy the new server mid-next week.

Stay tuned for more details on the changes on the search server!

UK mirror downtime

Our mirror server in the UK suffered a power supply failure yesterday. The mirror will be down for at least a week as the replacement power supply finds it way to the UK. We’ll post an update once the mirror is back online. Sorry for the hassle!

Our mirror server in the UK suffered a power supply failure yesterday. The mirror will be down for at least a week as the replacement power supply finds it way to the UK. We’ll post an update once the mirror is back online.

Sorry for the hassle!

Failed drive in database server RAID array (updated)

Yesterday afternoon a drive failed out of our RAID array in the master database server. We have four drives in this server, with two sets of two drives (using a RAID-1,0 configuration) so a single drive failure isn’t a big deal. We’re still up and running, but we’re quite a bit more vulnerable until we … Continue reading “Failed drive in database server RAID array (updated)”

Yesterday afternoon a drive failed out of our RAID array in the master database server. We have four drives in this server, with two sets of two drives (using a RAID-1,0 configuration) so a single drive failure isn’t a big deal. We’re still up and running, but we’re quite a bit more vulnerable until we get this drive replaced. If the last drive in our RAID-1 array fails we’d have to deal with data loss for today — fortunately the data dumps ran last night and backed us up until then. Let’s keep our fingers crossed that we do not have any failures today.

Once I got notified that the drive failed, I promptly called Sun for a warranty replacement. In the back of my mind I thought: “I should order a new drive from NewEgg, just in case”. But, the Sun representative assured my that they would have a replacement drive for me today. This morning a clueless service engineer called me to tell me that they shipped the wrong drive to our data center. When I called back, they informed me that they did not have the correct drive and that the vendor would ship one to me by Thursday morning.

Having lost faith in Sun to do the right thing, I ordered a spare drive for the server from NewEgg — I should’ve done this a couple of months ago when the prices for these drives become reasonable. Now we’ll have a spare on hand for the next time this happens (and it will quite soon, I’m sure — both Seagate drives that originally came with the server failed within one year. The first drive failed within the first month!)

So, I am curious to see who will win: Sun who had an 18 hour head start to replace the drive, or NewEgg/UPS who got my order just this morning. I’ll post an update once this plays out — in the meantime keep your fingers crossed for no further drive failures!

UPDATE: DHL delivered the Sun drives 1 hour before UPS delivered the NewEgg drives. The drive is now in the server and the RAID array is working optimally again.

Happy new year: MetaBrainz reaches sustainability!

Now that the holidays are over its time to get back to work and kick MusicBrainz work back into high gear! Its been my goal for many moons to make MetaBrainz self sufficient in order to not have to hold fundraisers in order to make ends meet. I’ve playing with these thoughts since 2002 when … Continue reading “Happy new year: MetaBrainz reaches sustainability!”

Now that the holidays are over its time to get back to work and kick MusicBrainz work back into high gear!

Its been my goal for many moons to make MetaBrainz self sufficient in order to not have to hold fundraisers in order to make ends meet. I’ve playing with these thoughts since 2002 when it looked like a business model could emerge with which to build a non-profit. And finally near the end of 2007 we reached this exciting milestone. Taking on new customers like Last.fm and the BBC made all the difference — we now have slightly more money coming in than we’re spending. We’re far from rich, but we’ve never been on more solid footing than this.

This gives me lots of hope for 2008! My plans now include trying to raise even more money so we can hire a full time server developer in order to make more progress on fixing bugs, writing new features and finally breaking ground on NGS. There is lots to do to keep things moving, but reaching sustainability is a strong vote of confidence from the universe that MusicBrainz and MetaBrainz may be up to something that could really work.

And that makes me smile. Onward!

P.S. Year end finances are more work to wrap up than usual. So, it will take longer than normal for me to post the finances for December and all of 2007.

Search server fixed.

UPDATE: The main search functions are back! FreeDB search will be offline until sometime tomorrow. Our search server finally gave up the ghost and its dead. We’re bringing up a back-up server to take over for it, but it will take another 8 hours for the indexes to be complete. Sorry for the inconvenience! In … Continue reading “Search server fixed.”

UPDATE: The main search functions are back! FreeDB search will be offline until sometime tomorrow.

Our search server finally gave up the ghost and its dead. We’re bringing up a back-up server to take over for it, but it will take another 8 hours for the indexes to be complete.

Sorry for the inconvenience!

In the meantime, you can donate some money!

UPDATE: by Dave at 1730 UTC

Until about 1715 UTC the search server was only available to the master site – this is now fixed. Sorry for the additional inconvenience, and thanks for bearing with us.

Previously Dave said:

The server server seems to be a little sick again today – the load average was over 100 just now. As a result I’m trying to reboot the server now, and I’ve put request rate-limiting in place on the main server, so you may find that various things aren’t working at the moment. 🙁

Watch this space for updates.

Two new editing suggestion reports available

Sergey and Wendell at MusicIP have been working on two new reports: Tracks attached to the wrong artist: This list contains tracks with similar names, which have been connected to artists which may be ambiguous – often this indicates an error in either the track assocation, or the artists themselves. To resolve the conflict, tracks … Continue reading “Two new editing suggestion reports available”

Sergey and Wendell at MusicIP have been working on two new reports:

  1. Tracks attached to the wrong artist: This list contains tracks with similar names, which have been connected to artists which may be ambiguous – often this indicates an error in either the track assocation, or the artists themselves. To resolve the conflict, tracks assigned to the incorrect artists should be moved to correct artists, or the artists should be merged if they represent the same musical source.
  2. PUIDs attached to the wrong artist: This list contains tracks which have the same PUID, but have been connected to conflicting artists. To resolve the conflict, PUIDs assigned incorrectly should be removed, tracks assigned to the incorrect artists should be moved to correct artists, or the artists should be merged if they represent the same musical source.

Thanks to Sergey and Wendell for creating these reports!

Last.fm contract signed!

Just a quick update on the contract with Last.fm: Last.fm has signed the contract with MusicBrainz and will now be able to integrate MusicBrainz data into their service once again. Three cheers! Thank you much to Spencer at Last.fm for pushing this through!

Just a quick update on the contract with Last.fm: Last.fm has signed the contract with MusicBrainz and will now be able to integrate MusicBrainz data into their service once again.

Three cheers!

Thank you much to Spencer at Last.fm for pushing this through!