Closed that1guy closed 9 years ago
The first link is a tad different - that will be dependent upon response times from AWS (which are at least 3 seconds).
Having said that, yes - these delays are a direct result of the posting-sync-agent. I killed all 3 (OR, WA, and CA) and reloaded both links above and they responded rapidly (the 2nd link was roughly 150 ms).
Looking at Jenkins and AWS console, it seems we're aggregating USA-CA again, thus causing this issue? If this is the case can we turn off ingest immediately and then start things back up around midnight tonight?
I just added a Jenkins job to help us with this moving forward...
In build.hashtagsell.com:8080 there is now a job named PAUSE-staging-posting-sync-agent.
Clicking the Build Now
link for this job will turn off the posting-sync-agents running against staging temporarily. The agents will be re-engaged if any of the following items occur:
Build Now
link is clicked for the job staging-posting-sync-agent-deploymentThanks for that. Sounds like this issue will be resolved by load balancer in prod env? Adding milestone to this issue.
Marking as resolved - see https://github.com/HashtagSell/posting-sync-agent/issues/11
Might this be happening because sync agent is hammering posting api?
Observed 9 sec - 44 sec (inconsistent)
Observed 4 sec - 36 seconds (inconsistent)