User:ApeKattQuest, MonkeyPython/aprilsnarr: Difference between revisions

From MusicBrainz Wiki
Jump to navigationJump to search
No edit summary
(reo's glorious rewrite)
Line 1: Line 1:
Our servers are on the move again!
We bring exciting news!
We all agreed that we will move our servers east again, this time all the way to India!


You might remember that a few years ago we moved all our servers from the US to Europe. Sadly, that has proven too costly* and prone to fires (https://www.datacenterdynamics.com/en/news/ovh-fire-update-four-halls-sbg1-destroyed-well-all-sbg2/), so we have decided our servers will continue their travel east and move to India, where hosting them will be significantly cheaper and cooling can be provided by sun-covering swarms of locusts. Most of the non-slacking members of our dev team are based in India by now anyway, so it feels like the right place!


As part of our cost-cutting process, and since air travel is not environmentally friendly anyway, we sent our servers East by boat. Unluckily, that means they had to wait their turn for quite a while in the only traffic jam you can see from space (https://www.esa.int/ESA_Multimedia/Images/2021/03/Suez_Canal_traffic_jam_seen_from_space). Luckily that is starting to get resolved now, but it will still take quite a while for the servers to clear the traffic jam, reach their destination and then be located among the twenty thousand containers in the ship carrying them, especially with all other jammed ships arriving at the same time! We're hoping to be done with it all in a couple months though, after which you can hopefully expect a return to our usual (not-particularly-)fast service.
[19:27] <yvanzo> Let's move our servers to the east again, maybe to India so we can just send them by ship (through Suez Canal.


In the meantime, our whole infrastructure is being hosted on some old servers that our friends at the Universitat Pompeu Fabra in Barcelona donated to us. They claimed they were getting too old for use in a rigorous academic environment, that as we all know in Spain requires systems to work continuously for a few hours every weekday with only a short break for a midday nap. These donated servers are now running from our executive director's basement. Any issues you have noticed recently are solely caused by this (and have nothing to do at all with the skill and attention span of our development team).
due to the rising cost of blah blah
we will again by moving our serverpark east this time to India
all of our non-slacker devs are already from india (except a few in spain, which is ironic)
bitmap doesn't count as he's so much west it becomes a rollower error and it counts as east again, much in accordance with the levels of productivity (east -> more productive)
due to the rising cost of green packaging we will be sending our servers through the newly UN-clogged suez channel
unfortunately our first shipment of servers where on ever last so that explains why ping time has been slow for a while.


* It could be argued that we should be able to cover the costs given we have had several years of good financial stability by now. While that is true, when we realized that the alternative is to cut costs and spend the savings on fancy ham and beers, the right choice became obvious to us.
because everyday is such a big boat it might also take some time for us to locate where the servers are stored - unpacking at our destination shouldn't take more than [ridiculous timeframe]

[the following was evelyns idea 👏 ]
to help in the meantime, we will be outsourcing some services to outer services in the meantime.
because things are so slow, we have found a temporary solution, outsourcing!

the relationships to the CDDB

classical relationships will be outsourced to Amazon Music HD
the new custom servers failed to arrive in time for the 21st anniversary of musicbrainz so Amazon kindly offered to help, as did cddb
we hope to celebrate with yet another cake ;)

listenbrainz data will be outsourced to last.fm but playlists will be outsourced to spotify
while bookbrainz is in its infancy so that not much working is status as usual

highly trained neural networks will be working for acousticbrainz for now, they are efficient and require just beer and the occasional sandwich.



evergood ever

Revision as of 15:27, 1 April 2021

Our servers are on the move again!

You might remember that a few years ago we moved all our servers from the US to Europe. Sadly, that has proven too costly* and prone to fires (https://www.datacenterdynamics.com/en/news/ovh-fire-update-four-halls-sbg1-destroyed-well-all-sbg2/), so we have decided our servers will continue their travel east and move to India, where hosting them will be significantly cheaper and cooling can be provided by sun-covering swarms of locusts. Most of the non-slacking members of our dev team are based in India by now anyway, so it feels like the right place!

As part of our cost-cutting process, and since air travel is not environmentally friendly anyway, we sent our servers East by boat. Unluckily, that means they had to wait their turn for quite a while in the only traffic jam you can see from space (https://www.esa.int/ESA_Multimedia/Images/2021/03/Suez_Canal_traffic_jam_seen_from_space). Luckily that is starting to get resolved now, but it will still take quite a while for the servers to clear the traffic jam, reach their destination and then be located among the twenty thousand containers in the ship carrying them, especially with all other jammed ships arriving at the same time! We're hoping to be done with it all in a couple months though, after which you can hopefully expect a return to our usual (not-particularly-)fast service.

In the meantime, our whole infrastructure is being hosted on some old servers that our friends at the Universitat Pompeu Fabra in Barcelona donated to us. They claimed they were getting too old for use in a rigorous academic environment, that as we all know in Spain requires systems to work continuously for a few hours every weekday with only a short break for a midday nap. These donated servers are now running from our executive director's basement. Any issues you have noticed recently are solely caused by this (and have nothing to do at all with the skill and attention span of our development team).

  • It could be argued that we should be able to cover the costs given we have had several years of good financial stability by now. While that is true, when we realized that the alternative is to cut costs and spend the savings on fancy ham and beers, the right choice became obvious to us.