#Buggy deployment since migration to EU West

23 messages · Page 1 of 1 (latest)

dry rock
#

Hello,

I just migrated our backend service (Django) + pg database to EU West.
The service is up, but:

  • it's slow compared to the staging service (that is hosted on US servers). The service is correctly setup and is using the private database address.
  • there is no deploy logs (cf screenshot)

Thanks for the help

winged bluffBOT
#

Project ID: 946c760b-c58f-4c9a-98b8-b3fe1d360c6c

dry rock
#

946c760b-c58f-4c9a-98b8-b3fe1d360c6c

dry rock
#

Any hint? Or should we move back our production to US servers? We won't be able to handle tomorrow load with those latencies.

solar turret
#

have you tried a simple redeploy?

dry rock
#

Yes, restart/redeploy/re-build image, does not work, I don't even see the echo from my docker entrypoint

solar turret
#

are you also getting no logs from postgres?

dry rock
#

I do get logs from postgres

solar turret
#

is postgres itself slow?

dry rock
#

No, seems ok

solar turret
#

okay and can you define slow in this case?

dry rock
solar turret
#

and what are those numbers when in us?

dry rock
#

(staging is hosted on US)

#

I'll try to get back to US for now

solar turret
#

are you located in EU?

dry rock
#

yes, created the service when EU servers were not available and we were trying to migrate to reduce latency

solar turret
#

is your service back in US now?

dry rock
#

Being migrated, there's a volume attached, so it's taking a bit

#

Back in Oregon, logs are working fine again and steady 30ms runtime on /ping.

solar turret
#

can you try to deploy a minimal reproducible example to EU?

dry rock
#

Will do before attempting to migrate again, but another day, it's getting late here.
Thanks for the help !