-
Notifications
You must be signed in to change notification settings - Fork 232
Senders fix, do not use static thrift factory #233
Senders fix, do not use static thrift factory #233
Conversation
Codecov Report
@@ Coverage Diff @@
## master #233 +/- ##
============================================
- Coverage 81.89% 81.87% -0.02%
+ Complexity 525 524 -1
============================================
Files 87 87
Lines 2005 2003 -2
Branches 236 236
============================================
- Hits 1642 1640 -2
Misses 263 263
Partials 100 100
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Change looks good. Do you think it’s worth having a test case with concurrent tracers?
We would have to emulate deserialization on the server size. Could we get it in and if necessary write test in separate PR? It's blocking my work |
Once this is merged. Could you please do micro release? (I don't have rights and @jpkrohling will be here on Wednesday) |
Yeah, we can write a test as a separate PR. I'll merge and release |
@pavolloffay are you saying the factory is not thread safe? |
I haven't digged very deep. I will write a test later and look more closely. |
The serializer isn't threadsafe, but looking at https://github.com/apache/thrift/blob/master/lib/java/src/org/apache/thrift/protocol/TBinaryProtocol.java#L85 it seems that we should be able to continue having the factory as static. |
When I was using multiple tracers in one process
HttpSender
was sending corrupted data.