Performance Impact of OpenResty XRay when Tracing Rust Apps
15 views
Skip to first unread message
Aisse Tejas
Jan 18, 2024, 12:00:27 PMJan 18
to openresty-en
Hello folks!
We recently published a new demo video https://youtu.be/bjmLEtwvIZI.
It shows how OpenResty XRay uses dynamic tracing technology to analyze the extra performance overhead introduced by a typical Rust target application using the `top` command.
The overhead of the OpenResty XRay analyzer itself and the impact on the target application’s performance is almost negligible. This makes it very suitable for online use, even for production environments with high service throughput and latency requirements.
We've put a lot of effort into reducing the overhead of OpenResty XRay itself. After all, many of our big customers are running it on production and are very sensitive to overhead. The most sensitive one is a cryptocurrency trading company, which is very sensitive to latency. They have a very large trading volume.
We always carefully check the sources of all extra latency and CPU overhead, while the open-source dynamic tracing toolchain usually does not care so much and often introduces very large latency or CPU loss here and there. Of course, sometimes we also contributed some of our patches to the open-source community, though that is just a small part of our many optimizations.
Text version: https://blog.openresty.com/en/rust-xray-overhead/
We recently published a new demo video https://youtu.be/bjmLEtwvIZI.
It shows how OpenResty XRay uses dynamic tracing technology to analyze the extra performance overhead introduced by a typical Rust target application using the `top` command.
The overhead of the OpenResty XRay analyzer itself and the impact on the target application’s performance is almost negligible. This makes it very suitable for online use, even for production environments with high service throughput and latency requirements.
We've put a lot of effort into reducing the overhead of OpenResty XRay itself. After all, many of our big customers are running it on production and are very sensitive to overhead. The most sensitive one is a cryptocurrency trading company, which is very sensitive to latency. They have a very large trading volume.
We always carefully check the sources of all extra latency and CPU overhead, while the open-source dynamic tracing toolchain usually does not care so much and often introduces very large latency or CPU loss here and there. Of course, sometimes we also contributed some of our patches to the open-source community, though that is just a small part of our many optimizations.
Text version: https://blog.openresty.com/en/rust-xray-overhead/
Reply all
Reply to author
Forward
0 new messages