https://linen.dev logo
Join Slack
Powered by
Hey everyone, I have starting moving things into ...
# troubleshooting
j
Hey everyone, I have starting moving things into production and now seeing an issues I had not when extracting from our production database
Copy code
'Instance <Job at 0x1069fb070> is not bound  │ │
│ │            to a Session; attribute refresh operation cannot proceed │ │
│ │            (Background on this error at:                            │ │
│ │            <https://sqlalche.me/e/14/bhk3)>'>                         │ │
│ │     self = <meltano.core.block.extract_load.
The database is a replicant running Postgress within Amazon RDS. The previous database were Postgres running in a docker container on my local laptop. I am still looking for database logs, but seems that the database session is being closed before the extractor finishes.
c
That's a really odd error message ... Following that SQL Alchemy link, it refers to the SQLAlchemy "lazy loading" technique, which is mainly used in the ORM features of SQLAlchemy. I'm a bit surprised that the Postgres Tap is making use of this SQLAlchemy technique as well. I'm assuming you are using the MeltanoLabs Tap-postgres variant.
Regardless, I do agree that you are likely running into a timeout issue on your read replica. My first guess would be that RDS Read Replicas are implemented as Hot Standby replicas. If permissible by the business requirements of your read replica (in case the read replica is shared with other business workloads apart from extract workload), I suggest you configure the 
max_standby_archive_delay
on the read replica in order to allow for long running SELECT statements.
j
yea, that is what we are looking at. WE set the 
max_standby_archive_delay
and 
max_standby_streaming_delay (integer)
to 3 hours and have not seen error in running a full sync
Open in Slack
PreviousNext
Powered by