You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is not a security vulnerability or a crashing bug
This is not a question about how to use Istio
Bug Description
We use Istio for both service mesh and as an ingress gateway.
We have a load test, using gatling, in open model. The load test sends a requests towards one of the services trough an ingress gateway:
load test --> ingress gateway --> upstream service
During the test, the upstream service fails, and ingress gateway starts to return 503, which is a correct behavior. What isn't correct, though, is the fact that ingress gateway memory starts to increase when the upstream service is down.
I would like to know what is the root cause of this issue, and why the memory starts to increase. This doesn't seem to be a memory leak, because when the test finishes, the memory is lowered significantly.
Below you can find graphs:
the first one is number of requests that ended up with either 200 or 503 code, this is based on istio_requests_total
the second one is memory reported by pod, based on container_memory_working_set_bytes
the third one is a p95 latency, measured by the ingress gateway pod, using istio_request_duration_milliseconds_bucket metric. The latency is measured by response code.
And lastly, a heap dump from the gateway, took during the time when memory usage was very high:
Is this the right place to submit this?
Bug Description
We use Istio for both service mesh and as an ingress gateway.
We have a load test, using gatling, in open model. The load test sends a requests towards one of the services trough an ingress gateway:
During the test, the upstream service fails, and ingress gateway starts to return 503, which is a correct behavior. What isn't correct, though, is the fact that ingress gateway memory starts to increase when the upstream service is down.
I would like to know what is the root cause of this issue, and why the memory starts to increase. This doesn't seem to be a memory leak, because when the test finishes, the memory is lowered significantly.
Below you can find graphs:
istio_requests_total
container_memory_working_set_bytes
istio_request_duration_milliseconds_bucket
metric. The latency is measured by response code.And lastly, a heap dump from the gateway, took during the time when memory usage was very high:
There are two questions that I would like to answer:
Version
Additional Information
No response
The text was updated successfully, but these errors were encountered: