|
i just lambdas and they all have names like buttstack273883DoThinglamba-prod-deadb33f-db9-83939494squishy what could be simpler
|
# ? May 21, 2023 04:09 |
|
|
# ? May 3, 2024 19:16 |
|
malificient_wombat
|
# ? May 22, 2023 17:44 |
|
we had PRD1 and PRD2 but they were on the same actual server so it was never clear why this was done also DR used to be (and by 'used to' I mean until mid 2022) manually moving the disk mounts to the other server and moving the CNAME lmao replication what's that???
|
# ? May 25, 2023 15:02 |
|
Powerful Two-Hander posted:also DR used to be (and by 'used to' I mean until mid 2022) manually moving the disk mounts to the other server and moving the CNAME lmao replication what's that??? DR == (Disk/DNS) Reconfiguration
|
# ? May 25, 2023 15:49 |
|
currently migrating some stuff from devops01 to devops03 so we can permanently decom devops01
|
# ? May 26, 2023 15:03 |
|
also we recently migrated from prodweb01 to prodweb10, prodweb02 to prodweb11, prodweb03 to prodweb12 etc
|
# ? May 26, 2023 15:05 |
|
Powerful Two-Hander posted:also DR used to be (and by 'used to' I mean until mid 2022) manually moving the disk mounts to the other server and moving the CNAME lmao replication what's that??? we do this, Snap Mirror to replicate the disks to the other DC and to swap you have to break the mirror, mount the LUNs onto the other machine, attach all the databases and change the DNS
|
# ? May 29, 2023 09:06 |
|
~Coxy posted:we do this, Snap Mirror to replicate the disks to the other DC and to swap you have to break the mirror, mount the LUNs onto the other machine, attach all the databases and change the DNS does it suck poo poo for you too? DR tests had to be scheduled and coordinated 3 months in advance because there were insufficient DBA resources to support more than a handful of servers failing over at once "So how would we actually do large scale fail over in the event of a disaster?" I asked "stop being a pain in the rear end and do the test " was the answer
|
# ? May 29, 2023 09:45 |
|
lol, I never thought of that (and I suspect neither did anybody else) it was probably designed more around the boring use case of single physical hardware failure so if Bentley DC burned down we would have been hosed, yeah
|
# ? May 29, 2023 10:20 |
|
it also took like 2 hours to do it lol when I had to fill in the architecture stuff I'd write "active/active (not the database)" and people got in a huff about that as well
|
# ? May 29, 2023 12:13 |
|
“do a dr test!” … “noooo! not that kind of dr!!”
|
# ? May 29, 2023 13:17 |
|
Continuous integration straight to prod.
|
# ? May 29, 2023 14:11 |
|
oh drat there were errors in the annual disaster recovery declaration, better fix it:quote:High availability? No, downtime! there we go, you can send that to the auditors now
|
# ? May 29, 2023 15:15 |
|
dev my prod hole
|
# ? May 29, 2023 19:31 |
|
naming servers like elon names his kids
|
# ? May 29, 2023 20:10 |
|
|
# ? Jun 17, 2023 13:59 |
|
Stick with the prod. Prod with the prod.
|
# ? Jun 17, 2023 15:03 |
|
renaming one of the prod servers Jessica
|
# ? Jun 17, 2023 15:51 |
|
|
# ? May 3, 2024 19:16 |
|
The Management posted:renaming one of the prod servers Jessica Crystal, Jessica, Amber
|
# ? Jun 17, 2023 17:43 |