Restart job after Redis data lose

Questions
,
1

Hi,

I have deploy N8N in queue mode (aka: one N8N-Main, a Redis DB and several N8N-Workers) and I’m testing the resolution of incidents/crashs.

I was testing data lose of the Redis, the use-case:

  1. Several jobs are in the Redis (ex: 1 in progress, 4 waiting)
  2. The Redis crash and I lost all the data
  3. Redis restart with empty database

In the PostgreSQL (table execution_entity), for these job/executions, the column finished is False and stoppedAt is empty.

Do I have a way/command to restart these jobs ?

Thx for the help

2

Hey @sGendrot,

If they appear under the exection log you should be good to go but if they have not started fully I am not sure if there is a way to handle it.

I know it doesn’t answer the question but how likely is it that you would have a failure with the Redis storage?

3

Hey @krynble, are we handling this scenario in any way or do you see an easy way to do so?

4

I can have a failure of the Redis, like I can have a failure of PostgreSQL. An application or someone can do a bad manipulation of the Redis and destroy data.

I already see both.