While investigating a brief viahtml outage, I noticed that some urllib requests were allowed to attempt to connect for > 2 minutes before timing out:
Long running requests like this tie up workers, so we should consider setting a shorter timeout on them (say 20 seconds or something).
While investigating a brief viahtml outage, I noticed that some urllib requests were allowed to attempt to connect for > 2 minutes before timing out:
Long running requests like this tie up workers, so we should consider setting a shorter timeout on them (say 20 seconds or something).