You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomcat.apache.org by bu...@apache.org on 2016/06/15 11:49:15 UTC

[Bug 59706] New: HTTP/2 load testing performance

https://bz.apache.org/bugzilla/show_bug.cgi?id=59706

            Bug ID: 59706
           Summary: HTTP/2 load testing performance
           Product: Tomcat 9
           Version: unspecified
          Hardware: All
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: Connectors
          Assignee: dev@tomcat.apache.org
          Reporter: remm@apache.org

Using a load testing tool like h2load, performance can go down quickly due to
very high multiplexing that occurs when processing the requests. Due to heavy
sync contention, it is likely some resources are then wasted. It could be
beneficial to explore solutions to that.

Example: h2load -c 1 -n 100 http://127.0.0.1:8080/tomcat.gif
Unlike HTTP/1.1, this will actually in most cases multiplex 100 concurrent
requests over a single connection, using 100 processing threads.

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


[Bug 59706] HTTP/2 load testing performance

Posted by bu...@apache.org.
https://bz.apache.org/bugzilla/show_bug.cgi?id=59706

--- Comment #3 from Remy Maucherat <re...@apache.org> ---
r1749898. I will patch 8.5 as well if there are no complaints.

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


[Bug 59706] HTTP/2 load testing performance

Posted by bu...@apache.org.
https://bz.apache.org/bugzilla/show_bug.cgi?id=59706

--- Comment #1 from Remy Maucherat <re...@apache.org> ---
Created attachment 33951
  --> https://bz.apache.org/bugzilla/attachment.cgi?id=33951&action=edit
Test code

Test code which hijacks activeRemoteStreamCount. I tried putting code in
StreamProcessor, among other possible options, but for some reason the
performance improvement is not as large.

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


[Bug 59706] HTTP/2 load testing performance

Posted by bu...@apache.org.
https://bz.apache.org/bugzilla/show_bug.cgi?id=59706

Remy Maucherat <re...@apache.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #4 from Remy Maucherat <re...@apache.org> ---
There's now a good balance between supporting HTTP/2 and limiting concurrency
enough.

-- 
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


[Bug 59706] HTTP/2 load testing performance

Posted by bu...@apache.org.
https://bz.apache.org/bugzilla/show_bug.cgi?id=59706

Remy Maucherat <re...@apache.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #33951|0                           |1
        is obsolete|                            |

--- Comment #2 from Remy Maucherat <re...@apache.org> ---
Created attachment 33956
  --> https://bz.apache.org/bugzilla/attachment.cgi?id=33956&action=edit
Queue v2

Same idea, revised.

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org