Category: NGS

As soon as we let people lose on the final test server with live data, people started finding lots more bugs. Which means that the final switchover will be delayed for a while longer.

But, the NGS data is live and replication packets for the NGS data are flowing. I’ll post instructions on how to tap into this feed later today.

UPDATE: Please be careful with your edits — we plan to keep this data!. Unless some data corruption bug is discovered, we’re going to do our best to keep this data. Please keep this in mind as you edit on this server.

We’re currently doing the last round of sanity checking to make sure we’ve not screwed anything up. Come help us test!

NOTE: This is our release candidate data set — please be careful what you edit. We’re strongly hoping to release with this data set!

If all goes well, expect us to go live at 9am BST on Tuesday 17 May!

We’re currently at 7 million edits migrated. We still have a lot of things to take care of, but we’re hoping to let people start playing with the migrated data at about 2:00am BST. We hope to go live at 9:00am BST.

See our updated schedule and maybe follow our updates on twitter.

We’re excited and we hope that you are too!

We had to fix some last minute migration issues and are running a test migration right now. If everything turns out well, we’ll be 8 hours behind schedule.

And after being more than a year late, what is 8 hours (or so…)?

The count of open edits has been dropping nicely and we’ve decided to push the no-edit date back to May 12. We could’ve done the 13th, but Friday the 13th didn’t seem like a good idea for this. 🙂

For more information on this, please read our preparing for NGS guide.

Once we move to NGS on May 16th, we’re going to require each application that makes requests to our web service to have a proper User-Agent header string. The User-Agent string needs to identify the application and the version of the application that is making the request; having a generic User-Agent string like “Java/1.6.0_24” or “PHP/5.3.4” does not allow us to properly identify the application making the requests.

IMPORTANT: 6 Months after we release NGS (Nov 16th) we’re going to start blocking common generic User-Agents strings, so please make sure that you send us a proper User-Agent header as part of your request.

We have a new NGS dataset available. This data set contains no known bugs (at this point in time) and is the basis for our restarted replication stream.

If you’re interesting in testing the NGS replication stream, please download and install this data set and then run admin/replication/LoadReplicationChanges to pull in the latest changes.

I’ve also uploaded a new NGS Virtual Machine — for details on how to use this virtual machine, please see the wiki page for the virtual machine. This VM is also setup to start replicating off this NGS data stream. Invoke the replication mechanism as mentioned above.

At this point we’re not certain if this is the last time we’re going to start the replication or if we’re going to do it one more time before we release NGS. Stay tuned!

We’ve started our reduced editing phase, where everyone is encouraged to not add any new edits to MusicBrainz until we release NGS. We’ve added ugly banners to each edit submission page and the home page on musicbrainz.org to remind everyone that they should be voting more than editing this week.

Next week we’re going to disallow edits entirely. For more information on these pre-NGS changes, please read our Preparing for NGS guide.

Lets get the edit queue down as fast as we can!

Murdos recently added a new report that highlights releases that have disc ids that will be discarded during the NGS data migration phase because the data is not conclusive enough to be reliably migrated. If you have some time, please help us review this data so we can save as much information when we switch over to NGS.

Thanks for creating that report Murdos!

(After fixing one buggy dataset, we have a couple of confirmations that the updated data set is ready to go.)

We have a new NGS dataset available — this data set still has a couple of migration bugs for ARs. The “RecordingA is cover of RecordingB” relationships are borked, but the rest of the data looks to be OK.

This new data also contains one very minor schema change: In the artist_credit_name table, the join_phrase column went from VARCHAR(32) to TEXT, to allow longer join phrases. This shouldn’t really affect anyone, but it is something to be aware of if you’re expecting a VARCHAR(32).

We’re going to work on fixing this bug and then we will put out a new data set and start replication off the new servers. In the mean time, feel free to play with the new data.

I’ve also uploaded a new NGS Virtual Machine — for details on how to use this virtual machine, please see the wiki page for the virtual machine.

UPDATE 1: Fixed the dataset and updated the link.
UPDATE 2: Added a note about the schema change.