Ensure Slack users are informed of timeouts
Ensure most cronjobs run in the peered VPCs
I cheekily slipped this in to this MR, but it's not related to the title
Ensure Slack users are informed of timeouts
If we use the same context to send slack messages that has just timed out trying to reach an external system, we won't be able to let the user know about the failure.
It's relatively harmless to ignore context for these updates. The global shared HTTP client has a 30s timeout configured already, and our Slack messages do not retry.
Also shorten the global async job timeout from 5 minutes to 2. 5 minutes is unrealistically long to expect a user to still expect updates!