Closed kreynoldsf5 closed 3 years ago
Packet captured locally:
Not a DNS issue The delay happens after the TLS handshake.
We'll need to debug within the app.
Suspended the loadgen job. I was able to repro the issue after several requests.
This means it is not an 'ambient load' issue from loadgen (although the additional load may make the delay happen more often).
Including the adservice in the deployment.
This pushed resource usage above thresholds for some REs. I re-distributed the loadgen to the 'state' vsite (which now needs to be renamed).
Average response time is WAY lower. Longest single request in the past hour is 2.1 seconds (entering the ADN in EMEA).
I will continue to monitor but this looks promising.
Intermittently, the response latency for a request will be disproportionately long (> 10s). Volterra dashboards suggests the latency is in the app (as opposed to ADN hops).
Investigate why this is happening.