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
Add a new event similar to dotnet/razor/taghelperresolve/end but is close to when Razor has completed discovering all of the razor components in a project,
Background and Motivation
During Automated Testing, when Visual Studio loads a solution, the LSP is running analysis in the background async. However, if the test infrastructure is trying to prevent regression of memory usage, things that happen asynchronously make it difficult to have a stable memory usage trace.
Right now, the tests are using Thread.Sleep(seconds) which is not the best way to capture regressions that may happen and not reliable.
…ty (#11233)
I was musing the other day about ETW events possibly being useful in
tracking down the Speedometer regression I was looking in to, and then
along comes #11232 which is a
great excuse to add some ETW event infrastructure to our lives.
The idea here is that RPS/Speedometer tests can wait for Razor to be
"ready" by looking for this event, rather than `Thread.Sleep`ing.
Anecdotally on my machine, when loading OrchardCore, this event fires
once a couple of minutes after solution load, so seems like a reasonable
candidate.
Once this is in @WardenGnaw will try it out and if it's not good, we'll
pop some more events somewhere.
Will need to follow up with PR(s?) to add this event source name to the
list, but I need to work out where to do that first :)
Summary
Add a new event similar to
dotnet/razor/taghelperresolve/end
but is close to when Razor has completed discovering all of the razor components in a project,Background and Motivation
During Automated Testing, when Visual Studio loads a solution, the LSP is running analysis in the background async. However, if the test infrastructure is trying to prevent regression of memory usage, things that happen asynchronously make it difficult to have a stable memory usage trace.
Right now, the tests are using
Thread.Sleep(seconds)
which is not the best way to capture regressions that may happen and not reliable.This used to be done by #3110
Proposed Feature
Add a new event like
dotnet/razor/taghelperresolve/end
or a CodeMarker .Alternative Designs
Some service that we can call that will wait when Razor is ready.
The text was updated successfully, but these errors were encountered: