Trustin Lee
b2d624a3ba
Warn if boss/worker threads are not up within 10 seconds
...
.. because it is very likely to be a user mistake.
- Fixes #1304
2013-10-17 21:12:03 +09:00
Trustin Lee
cc9c7d1607
Bump the version number to 3.8.0.Final
2013-10-17 20:31:19 +09:00
Jeff Pinner
75b52ac334
SPDY: remove SPDY/2 support
2013-10-17 20:29:05 +09:00
Trustin Lee
525dad5d01
Fix the problem where HashedWheelTimer puts a timeout into an incorrect place
...
- the stopIndex of a timeout is calculated based on the start time of the worker thread and the current tick count for greater accuracy
2013-10-07 17:17:12 +09:00
Jeff Pinner
49951b27c1
Merge pull request #1871 from jpinner/fix_write_stream_id
...
SPDY: create partial data frames using writeStreamId
2013-09-27 10:24:26 -07:00
Jeff Pinner
c96f732c9b
SPDY: create partial data frames using writeStreamId
2013-09-27 10:12:36 -07:00
Jeff Pinner
a0a6c9ed26
Merge pull request #1867 from jpinner/netty_3_spdy_3_1
...
SPDY/3.1
2013-09-27 07:45:11 -07:00
Jeff Pinner
bd54b516b7
SPDY: add SPDY/3.1 support
...
- with Michael Schore <mschore@twitter.com>
2013-09-26 21:48:58 -07:00
Jeff Pinner
fec0987cca
Merge pull request #1866 from jpinner/spdy_uni_dir_settings
...
SPDY: make MAX_CONCURRENT_STREAMS unidirectional
2013-09-26 09:57:46 -07:00
Jeff Pinner
dcced96bee
SPDY: make MAX_CONCURRENT_STREAMS unidirectional
2013-09-25 15:10:15 -07:00
Yuta Okamoto
5a44162c52
Fixes bugs when the backed array's offset is not zero.
2013-09-21 20:19:50 +02:00
Norman Maurer
ca2ee88d19
[ #1833 ] Replace ; with & only in the QueryString and not the whole URI. Also add test-case
2013-09-14 12:28:31 +02:00
Trustin Lee
a2532de2c7
Fix build error
2013-09-13 14:07:57 -07:00
Trustin Lee
34a21d3c35
Add the hello world HTTP server example to compare the performance against Netty 4
2013-09-13 13:53:43 -07:00
bgallagher
a512c099ca
fail future before firing exception event
2013-09-09 12:37:26 -07:00
Trustin Lee
df97732483
[maven-release-plugin] prepare for next development iteration
2013-09-05 23:14:49 +09:00
Trustin Lee
5b2cb54bb6
[maven-release-plugin] prepare release netty-3.7.0.Final
2013-09-05 23:14:39 +09:00
Trustin Lee
05ffb0bbe5
Ensure that SelectorUtil is initialized when AbstractNioSelector is initialized
...
- Related issue: #1701
- The dead lock reported in #1701 is basically because SelectorUtil's logger is initialized in a different thread, so I hope this will fix this specific issue.
2013-09-05 17:38:47 +09:00
Mike Schore
1f6a0407b6
SPDY: allow MAX_CONCURRENT_STREAMS to be set to 0
2013-09-05 16:45:36 +09:00
Jeff Pinner
bbf78a3434
Merge pull request #1745 from jpinner/fix_spdy_close_3
...
SPDY: ensure channel close happens after outbound message is written
2013-08-15 14:49:20 -07:00
Jeff Pinner
fd5d75de14
SPDY: ensure channel close happens after outbound message is written
2013-08-15 10:45:38 -07:00
bgallagher
eb80b4e104
intermittent: local channel sometimes sends messages after close
2013-08-09 21:11:17 +02:00
Norman Maurer
0cfbb7d54b
[ #1503 ]Fix a bug which incorrectly canceled keys for DatagramPackets when the epoll workaround was used
2013-07-22 11:47:02 +02:00
alain
5aeb836d8e
Proposal to improve performance of HttpResponseEncoder#encodeInitialLine
2013-07-11 16:29:30 +02:00
alain
240641a393
Improve performance of encodeHeader
2013-07-02 15:50:36 +09:00
Jeff Pinner
af59aa6ccb
SPDY: refactor SpdyHttpEncoder
2013-06-30 12:16:55 +09:00
Jeff Pinner
91ee0e5a56
SPDY: SpdyHeaderBlockDecoder should not discard input data
2013-06-29 09:42:21 +09:00
Jeff Pinner
b3ace85773
SPDY: SpdyFrameDecoder should buffer unread data when decoding headers
2013-06-28 21:28:54 +09:00
Trustin Lee
7654d3df02
Add missing license header / cleanup
2013-06-25 11:08:33 +09:00
Jeff Hodges
55360d2bc8
SPDY: handle too large header blocks
2013-06-25 11:08:33 +09:00
Trustin Lee
6aa4d147d8
Adds port to the host header value in WebSocket client handshake.
...
Due to http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.23 the port should be added if it differs from the default port. To simplify the code we just always add the port.
Backported from 82176bed76
2013-06-25 11:08:33 +09:00
Jeff Pinner
0ec12dcc60
SPDY: replace exception in SpdyHeaderBlockZlibEncoder with EMPTY_BUFFER
2013-06-25 11:08:33 +09:00
Trustin Lee
e0aabbddf2
Next version will be 3.7.0.Final
2013-06-25 11:08:33 +09:00
Mike Schore
e04df90698
SPDY: better encapsulation of header encoding/decoding
2013-06-25 11:08:26 +09:00
Jeff Pinner
e279803587
SPDY: update object hierarchy
2013-06-18 07:29:48 +09:00
Jeff Pinner
0b4adc0889
SPDY: always enqueue data frames if stream is stalled
2013-06-05 08:52:56 +02:00
Norman Maurer
6e1c5926ee
[ #1384 ] Cache HttpHeaderDateFormat in ThreadLocal
2013-05-27 06:57:15 +02:00
Norman Maurer
97332f6436
[ #1383 ] [ #1381 ] Encode / Decode the cookie expire data the same on all jvms
2013-05-22 14:01:38 +02:00
Norman Maurer
ca51aabc41
[ #1374 ] Allow users to all all subprotocols by using * as wildcard
2013-05-17 22:06:22 +02:00
Norman Maurer
7364317a01
[maven-release-plugin] prepare for next development iteration
2013-05-15 08:04:54 +02:00
Norman Maurer
90e1eb23ea
[maven-release-plugin] prepare release netty-3.6.6.Final
2013-05-15 08:04:43 +02:00
Norman Maurer
fd75994b16
[ #1358 ] Fix Encoding of multipart attribute names and filenames for non ASCII chars
2013-05-14 15:44:25 +02:00
Norman Maurer
1e380851bd
[ #1299 ] Make sure a handshake error will not trigger fireExceptionCaught two times
2013-05-12 13:36:38 +02:00
Norman Maurer
56c34214f8
[ #1338 ] Respect default ByteOrder of ChannelBufferFactory when using ProtoBufEncoder
2013-05-12 13:10:49 +02:00
Norman Maurer
23a919cdcd
[ #1007 ] HttpObjectAggregator should only throw one TooLongFrameException per full HTTP message
2013-05-09 20:03:02 +02:00
Norman Maurer
985b450182
Correctly handle multipart disposition
2013-05-08 21:59:00 +02:00
Frederic Bregier
1615009f5a
[ #1313 ] Fix for Multipart name and value that should not be encoded, except charset if specified
2013-05-08 10:29:28 +02:00
Norman Maurer
60695012ab
[ #1310 ] Fix deadlock which can happen if limit of MemoryAwareThreadPoolExecutor was exceed and an exceptionCaught(..) is triggered by a write
2013-05-06 13:54:25 +02:00
Norman Maurer
960067c78a
[ #1310 ] Fix deadlock which can happen if limit of MemoryAwareThreadPoolExecutor was exceed and a close is triggered by a write
...
* Close must be handled from outside the write lock to fix a possible deadlock
which can happen when MemoryAwareThreadPoolExecutor is used and the limit is exceed
and a close is triggered while the lock is hold. This is because the close(..)
may try to submit a task to handle it via the ExecutorHandler which then deadlocks.
2013-05-06 12:45:13 +02:00
Frederic Bregier
3371f4693b
Fix for #1313 & #1332
...
For issue #1313 , partial fix: enhance coding/deconding to missing part.
Now filename is correctly undecoded. Name and Value must be encoded to
allow parsing.
For issue #1332 , add text/plain to content-type as default type if not
specified.
2013-05-06 07:17:26 +02:00
