Replies: 1 comment 2 replies
-
Hi Luca, this is very interesting, thanks for the detailed analysis! Which version of Membrane are you using? Given this, I am confident we can reproduce and hopefully fix this. Best |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello,
I am using Membrane as a pass-through with an HTTPS and an HTTP target, using a proxies.xml file like the following.
When I call the HTTPS target directly, compared to calling it through Membrane, I observe an average difference of 120 milliseconds.
However, if the target URL uses the HTTP protocol instead of HTTPS, this difference is not noticeable.
While analyzing the httpClient interceptor, I noticed that enabling HTTP/2 (useExperimentalHttp2=true) allows the use of an httpClient pool, which in some cases reduces latency. However, this setup leads to system instability, where some requests take more than one second to execute, and I often receive com.predic8.membrane.core.util.EndOfStreamException.
On the other hand, if I do not use HTTP/2 (useExperimentalHttp2=false), there is no httpClientPool, and a new connection must be created for each request, which results in the observed 120 ms latency.
I would like to know if you are also experiencing this issue and if it is possible to configure an httpClientPool with useExperimentalHttp2=false to reduce this delay.
Thanks in advice.
Beta Was this translation helpful? Give feedback.
All reactions