Github user jeking3 commented on the issue:
https://github.com/apache/thrift/pull/1093
So overall, proper sequence ID handling is important. Most languages don't
need it for proper operation, as I suggested with C++ which processes one
request per connection at a time. In some work I did for Dell EqualLogic a
number of years ago, I leveraged a sequence number to rewrite the C# thrift
engine so it could support multiple outstanding requests, and effectively would
allow multiple threads to share the same connection and submit multiple
concurrent requests. So I do like to see proper sequence IDs in the
request/response for each language. It would be nice to have better tests
proving these work in two ways - one is to make sure that a seqId a client
passes in for each server language and protocol returns the seqId in the
response. The other would be to make sure every client actually sends a unique
seqId. Both are possible through crosstest with some work.
In the meantime though, having proper support for sequence IDs probably
starts in the protocol implementation. The protocol carries the sequence.
Generating a sequence ID in the compiler, or adding code that does seems
incorrect to me. It should be up to the constructor for a new binary protocol
header to generate a new sequence ID, and the server implementation should
honor it by carrying it through the response.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
