-
In your feedgen template repo's database.py file, what are the Why is updating the cursor or the database version necessary? I ask because I'm going to attempt to handle migrations in an offline Django app instead so I can manage additional new tables that way and do easy backup of a remote Postgres database, without having to figure out how to reimplement all of that in Flask. Then I just want to leave copies of the essential models in the Thanks for any insights. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
Saving the cursor to the database is required for data backfilling. For example, your server was down for 3 hours. And you want to continue from the moment when you stopped. You are getting a cursor from DB and continue work. There are plenty of scenarios including problems from the relay side (which could be an outage too). You should create an analog table using Django ORM which will keep cursors for firehose consumers. The meta table is the same as the Django migrations table. It was needed for migrating cursors between Firehouse URIs. For smooth migration from the old state of the repo. You can drop this table. |
Beta Was this translation helpful? Give feedback.
Saving the cursor to the database is required for data backfilling. For example, your server was down for 3 hours. And you want to continue from the moment when you stopped. You are getting a cursor from DB and continue work. There are plenty of scenarios including problems from the relay side (which could be an outage too). You should create an analog table using Django ORM which will keep cursors for firehose consumers.
The meta table is the same as the Django migrations table. It was needed for migrating cursors between Firehouse URIs. For smooth migration from the old state of the repo. You can drop this table.