Closed robertsmol closed 1 year ago
We identified the issue causing the unikernel instance to terminate, and fixed it in https://github.com/nanovms/nanos/pull/1847. When running an Envoy echo filter demo application on vSphere, I noticed that sometimes the server is abnormally slow to echo back the data it receives; I'm looking into why this is happening.
The low performance issue we observed with the Envoy echo filter is fixed in https://github.com/nanovms/nanos/pull/1848.
I am trying to get a sample application (an Envoy Filter) deployed in a Unikernel running on VSphere. My initial test for this deployment result in a Unikernel which eventually became unresponsive (it hung). After logging this issue the following fixes were made:
https://github.com/nanovms/lwip/commit/91c53301d6bf6ca4978a82f2cd6464c9ed18e7e4 https://github.com/nanovms/nanos/pull/1846
I tried out these fixes and the behavior has changed. The Unikernel now terminates. The following logs are produced:
boost_asio.txt logs_envoy.txt envoy_echo.txt