You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@trafficserver.apache.org by "Wendy Huang (JIRA)" <ji...@apache.org> on 2010/02/11 01:54:28 UTC
[jira] Created: (TS-164) 1)reducing timeout to 1ms; 2)using pipe fd
as signal between netthreads
1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
-----------------------------------------------------------------------
Key: TS-164
URL: https://issues.apache.org/jira/browse/TS-164
Project: Traffic Server
Issue Type: Improvement
Reporter: Wendy Huang
This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832837#action_12832837 ]
Leif Hedstrom commented on TS-164:
----------------------------------
John / Wendy: Is there any way we can "merge" the changes from dev branch that addresses this issue?
If not easily doable (i.e. we'd have to diverge trunk from dev even further), my +1 would be to postpone this "feature" until 2.1, when we merge dev branch into trunk.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "vijaya bhaskar mamidi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12836736#action_12836736 ]
vijaya bhaskar mamidi commented on TS-164:
------------------------------------------
Good find George. I looked at the code but i did not notice the mistake, looks like a copy and paste error.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch, TS164_patch2.diff
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "George Paul (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12835587#action_12835587 ]
George Paul commented on TS-164:
--------------------------------
Just looked at the patch. The way the iocore event and net data structures, methods etc have been re-worked/consolidated and used in the 'dev' branch to handle using eventfds & pipes for thread signaling makes this 'trunk' patch for the most part incompatible. When merging 'dev' back to 'trunk' we will have to resolve and address the incompatibilities then.
-George
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12836075#action_12836075 ]
Leif Hedstrom commented on TS-164:
----------------------------------
Fwiw, I tried with a completely fresh build (I even nuked the ccache), and same results, it hangs within a few minutes with this patch applied. Once it hangs, the only thing it does is
gettimeofday({1266626523, 917711}, NULL) = 0
epoll_wait(11, {}, 32768, 1) = 0
gettimeofday({1266626523, 918669}, NULL) = 0
epoll_wait(11, {}, 32768, 1) = 0
gettimeofday({1266626523, 919661}, NULL) = 0
epoll_wait(11, {}, 32768, 1) = 0
gettimeofday({1266626523, 920711}, NULL) = 0
epoll_wait(11, {}, 32768, 1) = 0
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Assigned: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "George Paul (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
George Paul reassigned TS-164:
------------------------------
Assignee: Leif Hedstrom
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (TS-164) 1)reducing timeout to 1ms; 2)using pipe fd
as signal between netthreads
Posted by "George Paul (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
George Paul updated TS-164:
---------------------------
Attachment: TS164_patch2.diff
A closer code review of original patch showed incorrect handling of EVENTFD type case in mainNetEvent handler (copy & paste error? ). This patch 'TS164_patch2.diff' fixes this and is a replacement for the original patch 'TS-164.patch' . Tested in forward proxy mode and no longer see TS 'hung' state.
-George
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch, TS164_patch2.diff
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12836110#action_12836110 ]
Leif Hedstrom commented on TS-164:
----------------------------------
If there is no resolution to this by Monday morning, I think we'll have to postpone this feature to 2.0.0 (I'm not sure I like that idea), or, wait for 2.2.0, which has an alternative solution already in place (on the dev branch).
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (TS-164) 1)reducing timeout to 1ms; 2)using pipe fd
as signal between netthreads
Posted by "Wendy Huang (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Wendy Huang updated TS-164:
---------------------------
Attachment: TS-164.patch
patch file attached.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Leif Hedstrom resolved TS-164.
------------------------------
Resolution: Fixed
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch, TS164_patch2.diff
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Work started: (TS-164) 1)reducing timeout to 1ms; 2)using
pipe fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Work on TS-164 started by Leif Hedstrom.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12836059#action_12836059 ]
Leif Hedstrom commented on TS-164:
----------------------------------
This is the stack thread from all threads when the traffic_server process "hangs" (it's no longer accepting new connections):
(gdb) thread apply all bt
Thread 20 (process 12336):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x404a4fc8, event=5, e=0xa19b9e0)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x404a4008, e=0xa19b9e0,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x404a4008) at UnixEThread.cc:249
#5 0x08109f89 in main (argc=5, argv=0xafb0624) at Main.cc:2096
Thread 19 (process 12356):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b65bbb in write () from /lib/libpthread.so.0
#2 0x082e3e7e in ProtectedQueue::signal (this=0x408a7aa0)
at P_UnixSocketManager.h:239
#3 0x082bc5df in NetAccept::do_blocking_accept (this=0xae69c58, master_na=0xae69c58,
t=0x44bb9008) at ../../iocore/eventsystem/P_UnixEventProcessor.h:80
#4 0x082bc9f2 in net_accept_main_blocking (na=0xae69c58, e=0xa19b260, blockable=true)
at UnixNetAccept.cc:199
#5 0x082bca6a in NetAccept::acceptLoopEvent (this=0xae69c58, event=1, e=0xa19b260)
at UnixNetAccept.cc:635
#6 0x082e5c96 in EThread::execute (this=0x44bb9008) at I_Continuation.h:147
#7 0x082e430c in spawn_thread_internal (a=0xae69cf0) at Thread.cc:85
#8 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#9 0x00ac520e in clone () from /lib/libc.so.6
Thread 18 (process 12355):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b65bbb in write () from /lib/libpthread.so.0
#2 0x082e3e7e in ProtectedQueue::signal (this=0x408a7aa0)
at P_UnixSocketManager.h:239
#3 0x082bc5df in NetAccept::do_blocking_accept (this=0xae60068, master_na=0xae60068,
t=0x44916008) at ../../iocore/eventsystem/P_UnixEventProcessor.h:80
#4 0x082bc9f2 in net_accept_main_blocking (na=0xae60068, e=0xa19b320, blockable=true)
at UnixNetAccept.cc:199
#5 0x082bca6a in NetAccept::acceptLoopEvent (this=0xae60068, event=1, e=0xa19b320)
at UnixNetAccept.cc:635
#6 0x082e5c96 in EThread::execute (this=0x44916008) at I_Continuation.h:147
#7 0x082e430c in spawn_thread_internal (a=0xae60030) at Thread.cc:85
#8 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#9 0x00ac520e in clone () from /lib/libc.so.6
Thread 17 (process 12353):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ed2 in pthread_cond_timedwait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x081d0bd9 in Log::flush_thread_main (args=0x0)
at ../../libinktomi++/ink_thread.h:253
#3 0x081d393d in LoggingFlushContinuation::mainEvent (this=0xae10448, event=1,
data=0xa19b420) at Log.cc:329
#4 0x082e5c96 in EThread::execute (this=0x44714008) at I_Continuation.h:147
#5 0x082e430c in spawn_thread_internal (a=0xa1fa3b8) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 16 (process 12352):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd27e in PollCont::pollEvent (this=0x44472400, event=5, e=0xa19b4a0)
at UnixNet.cc:84
#3 0x082e4bc1 in EThread::process_event (this=0x44471008, e=0xa19b4a0,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x44471008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa1e55d8) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 15 (process 12351):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ed2 in pthread_cond_timedwait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x082b2b67 in AIOThreadInfo::start (this=0xa1e4fc8, event=1, e=0xa19b4e0)
at ../../libinktomi++/ink_thread.h:253
#3 0x082e5c96 in EThread::execute (this=0x4426f008) at I_Continuation.h:147
#4 0x082e430c in spawn_thread_internal (a=0xa1e4ff0) at Thread.cc:85
#5 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#6 0x00ac520e in clone () from /lib/libc.so.6
Thread 14 (process 12350):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ba5 in pthread_cond_wait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x082b2b83 in AIOThreadInfo::start (this=0xa1e4ee8, event=1, e=0xa19b520)
at ../../libinktomi++/ink_thread.h:247
#3 0x082e5c96 in EThread::execute (this=0x4406d008) at I_Continuation.h:147
#4 0x082e430c in spawn_thread_internal (a=0xa1e4f10) at Thread.cc:85
#5 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#6 0x00ac520e in clone () from /lib/libc.so.6
Thread 13 (process 12349):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ba5 in pthread_cond_wait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x082b2b83 in AIOThreadInfo::start (this=0xa1e4e08, event=1, e=0xa19b560)
at ../../libinktomi++/ink_thread.h:247
#3 0x082e5c96 in EThread::execute (this=0x43e6b008) at I_Continuation.h:147
#4 0x082e430c in spawn_thread_internal (a=0xa1e4e30) at Thread.cc:85
#5 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#6 0x00ac520e in clone () from /lib/libc.so.6
Thread 12 (process 12348):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ba5 in pthread_cond_wait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x082b2b83 in AIOThreadInfo::start (this=0xa1e4d38, event=1, e=0xa19b5a0)
at ../../libinktomi++/ink_thread.h:247
#3 0x082e5c96 in EThread::execute (this=0x43c69008) at I_Continuation.h:147
#4 0x082e430c in spawn_thread_internal (a=0xa1dc5d0) at Thread.cc:85
#5 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#6 0x00ac520e in clone () from /lib/libc.so.6
Thread 11 (process 12347):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00a81736 in nanosleep () from /lib/libc.so.6
#2 0x00a8154e in __sleep (seconds=<value optimized out>)
at ../sysdeps/unix/sysv/linux/sleep.c:138
#3 0x082df29a in stat_sync_cont::sync (this=0xa195798, event=1, e=0xa19ba60)
at RecProcess.cc:262
#4 0x082e5c96 in EThread::execute (this=0x417b7008) at I_Continuation.h:147
#5 0x082e430c in spawn_thread_internal (a=0xa1abd68) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 10 (process 12346):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00a81736 in nanosleep () from /lib/libc.so.6
#2 0x00a8154e in __sleep (seconds=<value optimized out>)
at ../sysdeps/unix/sysv/linux/sleep.c:138
#3 0x082df3d3 in sync_cont::sync (this=0xa195840, event=1, e=0xa19baa0)
at RecProcess.cc:240
#4 0x082e5c96 in EThread::execute (this=0x415b5008) at I_Continuation.h:147
#5 0x082e430c in spawn_thread_internal (a=0xa195778) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 9 (process 12345):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00a81736 in nanosleep () from /lib/libc.so.6
#2 0x00a8154e in __sleep (seconds=<value optimized out>)
at ../sysdeps/unix/sysv/linux/sleep.c:138
#3 0x082df1e2 in config_update_cont::exec_callbacks (this=0xa195820, event=1,
e=0xa19bae0) at RecProcess.cc:201
#4 0x082e5c96 in EThread::execute (this=0x413b3008) at I_Continuation.h:147
#5 0x082e430c in spawn_thread_internal (a=0xa1948c0) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 8 (process 12344):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00a81736 in nanosleep () from /lib/libc.so.6
#2 0x00a8154e in __sleep (seconds=<value optimized out>)
at ../sysdeps/unix/sysv/linux/sleep.c:138
#3 0x082df10a in raw_stat_sync_cont::exec_callbacks (this=0xa1955b8, event=1,
e=0xa19bb20) at RecProcess.cc:179
#4 0x082e5c96 in EThread::execute (this=0x411b1008) at I_Continuation.h:147
#5 0x082e430c in spawn_thread_internal (a=0xa194888) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 7 (process 12343):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00b62ed2 in pthread_cond_timedwait@@GLIBC_2.3.2 () from /lib/libpthread.so.0
#2 0x082e3d44 in ProtectedQueue::dequeue_timed (this=0x410afaa0,
cur_time=1266625009463389000, timeout=1266625009523389000, sleep=true)
at ../../libinktomi++/ink_thread.h:253
#3 0x082e5af9 in EThread::execute (this=0x40faf008) at UnixEThread.cc:266
#4 0x082e430c in spawn_thread_internal (a=0xa133768) at Thread.cc:85
#5 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#6 0x00ac520e in clone () from /lib/libc.so.6
Thread 6 (process 12342):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x409a9fc8, event=5, e=0xa19b760)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x409a9008, e=0xa19b760,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x409a9008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa14ecf0) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 5 (process 12341):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x408a8fc8, event=5, e=0xa19b7e0)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x408a8008, e=0xa19b7e0,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x408a8008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa195880) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 4 (process 12340):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x407a7fc8, event=5, e=0xa19b860)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x407a7008, e=0xa19b860,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x407a7008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa1955e8) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 3 (process 12339):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x406a6fc8, event=5, e=0xa19b8e0)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x406a6008, e=0xa19b8e0,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x406a6008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa195480) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 2 (process 12338):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00ac59d6 in epoll_wait () from /lib/libc.so.6
#2 0x082bd7a0 in NetHandler::mainNetEvent (this=0x405a5fc8, event=5, e=0xa19b960)
at UnixNet.cc:221
#3 0x082e4bc1 in EThread::process_event (this=0x405a5008, e=0xa19b960,
calling_code=5) at I_Continuation.h:147
#4 0x082e5339 in EThread::execute (this=0x405a5008) at UnixEThread.cc:249
#5 0x082e430c in spawn_thread_internal (a=0xa194868) at Thread.cc:85
#6 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#7 0x00ac520e in clone () from /lib/libc.so.6
Thread 1 (process 12337):
#0 0x40000416 in __kernel_vsyscall ()
#1 0x00a81736 in nanosleep () from /lib/libc.so.6
#2 0x00a8154e in __sleep (seconds=<value optimized out>)
at ../sysdeps/unix/sysv/linux/sleep.c:138
#3 0x08227999 in startProcessManager (arg=0x0) at ProcessManager.cc:83
#4 0x00b5f32f in start_thread (arg=<value optimized out>) at pthread_create.c:297
#5 0x00ac520e in clone () from /lib/libc.so.6
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (TS-164) 1)reducing timeout to 1ms; 2)using pipe fd
as signal between netthreads
Posted by "George Paul (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
George Paul updated TS-164:
---------------------------
Component/s: Core
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Components: Core
> Reporter: Wendy Huang
> Assignee: Leif Hedstrom
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12835576#action_12835576 ]
Leif Hedstrom commented on TS-164:
----------------------------------
Wendy, can you check with the "dev" branch and John's changes, and see if what we got is "close enough" ? When done, assign the bug to me, and I'll review the (final) patch.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "Wendy Huang (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832674#action_12832674 ]
Wendy Huang commented on TS-164:
--------------------------------
ah. just saw john's comment, I'll take a look on what you mentioned.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "John Plevyak (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12833067#action_12833067 ]
John Plevyak commented on TS-164:
---------------------------------
When I talked to VJ it sounds like the patch is very similar. If we want it in trunk, VJ's patch is probably easier to incorporate.
I think we should make the decision independently.
I was just suggesting that when merging into trunk you might want to look at the dev line and note any differences
to make sure nothing is overlooked and likewise when the trunk change is merged into dev.
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (TS-164) 1)reducing timeout to 1ms; 2)using pipe fd
as signal between netthreads
Posted by "Leif Hedstrom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Leif Hedstrom updated TS-164:
-----------------------------
Fix Version/s: 2.0.0a
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
> Fix For: 2.0.0a
>
> Attachments: TS-164.patch
>
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (TS-164) 1)reducing timeout to 1ms; 2)using pipe
fd as signal between netthreads
Posted by "John Plevyak (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/TS-164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832640#action_12832640 ]
John Plevyak commented on TS-164:
---------------------------------
When doing the merge, please take a look a the dev branch
which has fixed this problem using eventfd when available with
fallback to pipe().
The bug was TS-94
http://issues.apache.org/jira/browse/TS-94
> 1)reducing timeout to 1ms; 2)using pipe fd as signal between netthreads
> -----------------------------------------------------------------------
>
> Key: TS-164
> URL: https://issues.apache.org/jira/browse/TS-164
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: Wendy Huang
>
> This is copied from bugzilla 3095908. two type of changes made in this bug as indicated in summary.
> Need to merge the change to open source.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.