Repository navigation
Stable 2.3.1 Release Notes - #2869
Conversation
There was a problem hiding this comment.
Looks good, and I just had super nit picky formatting comments, but will defer to the proxy folks to approve the content.
| * Fixed a bug in the proxy's HPACK codec that could cause requests with | ||
| very large header values to hang indefinitely. | ||
| * Replaced the fixed reconnect backoff with an exponential one (thanks, | ||
| @zaharidichev!) |
There was a problem hiding this comment.
You can remove 2 leading spaces from this line, to match to formatting of the other bullet points.
| queries to the Destination service | ||
| * Fix an epoll notification issue that could cause excessive CPU usage | ||
| * Added the ability to disable tap by setting an env var (thanks, | ||
| @zaharidichev!) |
| **Full release notes**: | ||
| * Proxy | ||
| * Changed the proxy's routing behavior so that, when the control plane | ||
| does not resolve a destination, the proxy forwards request with minimal |
There was a problem hiding this comment.
nice! 🚢 🥇 2 minor nits
| **Full release notes**: | ||
| * Proxy | ||
| * Changed the proxy's routing behavior so that, when the control plane | ||
| does not resolve a destination, the proxy forwards request with minimal |
There was a problem hiding this comment.
should this be proxy forwards the request?
| does not resolve a destination, the proxy forwards request with minimal | ||
| additional routing logic | ||
| * Fixed a bug in the proxy's HPACK codec that could cause requests with | ||
| very large header values to hang indefinitely. |
There was a problem hiding this comment.
you can remove period from the end of this sentence
| * Added a dispatch timeout that limits the amount of time a request can be | ||
| buffered in the proxy | ||
| * Removed the limit on the number of concurrently active service discovery | ||
| queries to the Destination service |
There was a problem hiding this comment.
total nit, but Capitalization of "Destination service" seems strange, maybe either both caps or neither? If that is indeed correct, feel free to ignore.
There was a problem hiding this comment.
I was curious about this too as I'm drafting edge release notes. Right now there are 3 "Destination service" mentions and 3 "destination service" mentions so Dennis I guess you are the tie-breaker? :)
There was a problem hiding this comment.
This is too much pressure haha. No caps feels much better?
|
Integration test results for 7b26b4b: success 🎉 |
|
Thanks for the feedback folks, changes coming up in the next commit. |
|
Integration test results for eb62737: success 🎉 |
|
|
||
| * Proxy | ||
| * Changed the proxy's routing behavior so that, when the control plane | ||
| does not resolve a destination, the proxy forwards the request with minimal |
There was a problem hiding this comment.
IMO "forwards the request to its original destination" better describes the behaviour than "forwards the request with minimal additional routing logic".
There was a problem hiding this comment.
I think "forwards the request with minimal additional routing logic" works because it implies there is more going on behind the scenes w.r.t routing and that the behavior has now been fixed.
| buffered in the proxy | ||
| * Removed the limit on the number of concurrently active service discovery | ||
| queries to the destination service | ||
| * Fix an epoll notification issue that could cause excessive CPU usage |
There was a problem hiding this comment.
This is in present tense, while it looks like all the other bullet points are in past-tense.
| * Fix an epoll notification issue that could cause excessive CPU usage | |
| * Fixed an epoll notification issue that could cause excessive CPU usage |
| very large header values to hang indefinitely | ||
| * Replaced the fixed reconnect backoff with an exponential one (thanks, | ||
| @zaharidichev!) | ||
| * Fixed an issue where load balancers can become stuck |
There was a problem hiding this comment.
Perhaps: Fixed an issue where requests could be held indefinitely by the load balancer ?
"stuck" is a bit squishy
There was a problem hiding this comment.
Sounds great. Thanks for the feedback.
|
Integration test results for d890edf: success 🎉 |
|
Integration test results for 2d9cfff: fail 😕 |
This stable release adds a number of proxy stability improvements. To install this release, run: `curl https://run.linkerd.io/install | sh` **Special thanks to**: @zaharidichev and @11Takanori! **Full release notes**: * Proxy * Changed the proxy's routing behavior so that, when the control plane does not resolve a destination, the proxy forwards the request with minimal additional routing logic * Fixed a bug in the proxy's HPACK codec that could cause requests with very large header values to hang indefinitely * Replaced the fixed reconnect backoff with an exponential one (thanks, @zaharidichev!) * Fixed an issue where requests could be held indefinitely by the load balancer * Added a dispatch timeout that limits the amount of time a request can be buffered in the proxy * Removed the limit on the number of concurrently active service discovery queries to the destination service * Fixed an epoll notification issue that could cause excessive CPU usage * Added the ability to disable tap by setting an env var (thanks, @zaharidichev!)
This stable release adds a number of proxy stability improvements.
To install this release, run:
curl https://run.linkerd.io/install | shSpecial thanks to: @zaharidichev and @11Takanori!
Full release notes:
does not resolve a destination, the proxy forwards the request with minimal
additional routing logic
very large header values to hang indefinitely
@zaharidichev!)
buffered in the proxy
queries to the destination service
@zaharidichev!)