Hello,
As of today, the SCB cluster consists of 4 machines in each DC, which spreads the load more evenly and allows for better handling of bad situations. Kudos to Alex who carried out the work of re-purposing the ex-SCA nodes (by previously moving Zotero to VMs)~[1]. The deploy repositories of SCB services have been updated~[2], so you don't have to adapt to the new situation - just deploy as normal, but see your service deployed to 8 instead of 4 nodes total :)
Cheers, Marko
[1] https://phabricator.wikimedia.org/T148380 [2] https://phabricator.wikimedia.org/T147903
This is great to have MCS and other node services running with twice the capacity. Thank you to all involved!
Cheers, Bernd
On Wed, Oct 19, 2016 at 11:47 AM, Marko Obrovac mobrovac@wikimedia.org wrote:
Hello,
As of today, the SCB cluster consists of 4 machines in each DC, which spreads the load more evenly and allows for better handling of bad situations. Kudos to Alex who carried out the work of re-purposing the ex-SCA nodes (by previously moving Zotero to VMs)~[1]. The deploy repositories of SCB services have been updated~[2], so you don't have to adapt to the new situation - just deploy as normal, but see your service deployed to 8 instead of 4 nodes total :)
Cheers, Marko
[1] https://phabricator.wikimedia.org/T148380 [2] https://phabricator.wikimedia.org/T147903
-- Marko Obrovac, PhD Senior Services Engineer Wikimedia Foundation
Services mailing list Services@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/services