GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/772
TS-3726: Expose build_error_response
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver TS-3726
Alternatively you can
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/771
Proposal: InactivityCop Optimize
Optimize InactivityCop on a heavy load and high level concurrent connection
system.
It is set default inactivity timeout before put vc into open_list
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/766
TS-4614: avoid e->schedule_in for dummy event
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver TS-4614
Alternativ
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/701
@shinrich I remember that SSL_write would return âSSL_WANT_READâ error
code.
from: https://www.openssl.org/docs/manmaster/ssl/SSL_write.html
If the underlying BIO is non
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/753
@shinrich the "noisy" in UnixNetAccept, let me explain, there are 3 area to
create a new netvc and init its members but disorder, I'm try to make it
ordered as below
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/753#discussion_r68782930
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -443,19 +430,28 @@ NetAccept::acceptFastEvent(int event, void *ep)
goto Lerror
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/752
@shinrich INKVConnInternal::handle_event is almost same as
INKContInternal::handle_event but the INKContInternal::handle_event could
handle period event callback status.
They are share
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/752#discussion_r68778438
--- Diff: proxy/InkAPI.cc ---
@@ -1037,43 +1047,36 @@ INKVConnInternal::INKVConnInternal(TSEventFunc
funcp, TSMutex mutexp)
: INKContInternal
Github user oknet closed the pull request at:
https://github.com/apache/trafficserver/pull/673
---
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
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/754
This is a rebase version of #673 .
---
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
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/673
@zwoop @shinrich a new PR#754 created based on the newest master branch.
---
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
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/754
TS-4487: check the change of lock after return from wbe callback
and handle (needs & EVENTIO_READ) first.
You can merge this pull request into a Git repository by running:
$ git
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/642
@zwoop a new PR#753 created based on the newest master branch.
---
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
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/753
This is a rebase version of #642
---
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
Github user oknet closed the pull request at:
https://github.com/apache/trafficserver/pull/642
---
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
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/753#discussion_r68736872
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -281,29 +280,28 @@ NetAccept::do_blocking_accept(EThread *t)
return -1
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/753#discussion_r68736720
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -281,29 +280,28 @@ NetAccept::do_blocking_accept(EThread *t)
return -1
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/753#discussion_r68736528
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -117,11 +117,10 @@ net_accept(NetAccept *na, void *ep, bool blockable)
vc->
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/642
@zwoop a new PR#753 created based on the newest master branch.
---
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
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/753
Proposal: NetVC Context
Goal 1st:
In the NetVConnection, we have get_local_addr() and get_remote_addr()
method.
Also have members local_addr, remote_addr and netvc->con.a
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/642#discussion_r68717045
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -277,14 +277,12 @@ NetAccept::do_blocking_accept(EThread *t)
vc->options.packet_mark = packet_m
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/642#discussion_r68715180
--- Diff: proxy/InkAPI.cc ---
@@ -6249,12 +6249,12 @@ TSVConnFdCreate(int fd)
vc->action_ = &a;
vc->id = net_next_connect
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/752
TS-4590: INKContInternal & INKVConnInternal Optimize
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver TS-
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/642
@zwoop I will create another PR and rebase on the newest master branch.
---
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
Github user oknet closed the pull request at:
https://github.com/apache/trafficserver/pull/665
---
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
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/665
@jpeach create a new PR#734 base on master.
---
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
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/734
TS-4483: NetAccept & SSLNetAccept optimize, replace getEtype() with mâ¦
â¦ember etype.
You can merge this pull request into a Git repository by running:
$ git pull https://github
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/718#discussion_r67817437
--- Diff: proxy/http/HttpSM.cc ---
@@ -2722,7 +2722,10 @@ HttpSM::tunnel_handler_post(int event, void *data)
{
STATE_ENTER(&Ht
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/718#discussion_r67816840
--- Diff: proxy/http/HttpSM.cc ---
@@ -2841,9 +2844,9 @@ HttpSM::tunnel_handler_push(int event, void *data)
ink_assert(data == &tu
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/701
I'm try do some test and get result below:
ECONNRESET from write() while get TCP_RST from peer.
EPIPE from write() while get TCP_FIN from peer.
SSLNetVConne
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/701
@jpeach ECONNRESET is only found in man 2 send/sendto/sendmsg. obviously
EPIPE for write() the same as ECONNRESET for send()/sendto()/sendmsg(). And
write() is called in load_buffer_and_write
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/701
@jpeach VC_EVENT_EOS indicates socket fd is closed. VC_EVENT_ERROR means
meet a error and can not going on. Looking around the comment on
I_VConnection.h about do_io_read and do_io_write
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/701
TS-4522: check EPIPE instead r==0
man 2 write, Return value, On Success, the number of bytes written is
returned (zero indicates nothing was written). On error, -1 is returned, and
errno is
Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/700
maybe should add EPIPE check?
```
if (!r || r == -ECONNRESET || r == -EPIPE) { // SIGPIPE when write on
a closed fd.
vc->write.triggered = 0;
write_signal_d
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/673#issuecomment-222517497
Due to the #629 already merged, I will update the patch to fit it.
---
If your project is set up for it, you can reply to this email and have your
reply appear on
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/673#issuecomment-222300709
vote on PR #629, but the 2nd issue: did not check the change of lock after
return from wbe callback, not included.
---
If your project is set up for it, you can
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/629#discussion_r64986856
--- Diff: iocore/net/SSLNetVConnection.cc ---
@@ -193,131 +193,121 @@ ssl_read_from_net(SSLNetVConnection *sslvc, EThread
*lthread, int64_t &
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/673
TS-4487: fix write_to_net_io issues
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patch-13
Alternatively you can
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/665#issuecomment-22179
@shinrich ET_SSL is still here. SSLNetAccept::etype is set to ET_SSL by
sslNetProcessor::upgradeEtype().
---
If your project is set up for it, you can reply to
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/665#issuecomment-221612012
Sorry, forget to create it, please link to TS-4483.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/665
NetAccept & SSLNetAccept Optimize, replace getEtype() with member etype.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/o
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/568#discussion_r63826235
--- Diff: lib/ts/Diags.cc ---
@@ -625,11 +625,14 @@ Diags::should_roll_diagslog()
fflush(diags_log->m_fp);
if (diags_log-&g
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/646
TS-4455: remove isEosRcvd() and eosRcvd
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patch-12
Alternatively you
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/642#issuecomment-219947645
review the code again, the variable name 'server_addr' confuse me. now, I'm
using get_remote_addr() to replace server_addr and remote get
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/642#issuecomment-219936052
@SolidWallOfCode get_server_addr() and get_client_addr() is wrapper
functions only.
the wrapper function get_server_addr() is used to decide which
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/642#issuecomment-219642314
Currently, get_server_addr() and get_client_addr() implement as wrapper
functions to call get_local_addr() and get_remote_addr() depend on netvc
context. It is
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/642#issuecomment-219613493
@SolidWallOfCode Let me explain it. In the Client<-->ATS side, client_addr
is Client IP Address and server_addr is ATS service IP Address. In the other
side
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/642
Proposal: NetVC Context
In the NetVConnection, we have get_local_addr() and get_remote_addr()
method.
Also have members local_addr, remote_addr and netvc->con.addr.
Thus,
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/625#issuecomment-218362265
int
UnixNetVConnection::connectUp(EThread *t, int fd)
{
int res;
thread = t;
if (check_net_throttle(CONNECT, submit_time
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/626
TS4433: Did not consider NULL return from TSVConnFdCreate
example plugin: intercept
Did not consider NULL return from TSVConnFdCreate only do a check on
connect.
You can merge this pull
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/625
TS-4432: TSVConnFdCreate, double free on connectUp failure
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patch-9
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/373#issuecomment-212809087
@shinrich I'm closed the PR, but github show me "but the oknet:patch-1
branch has unmerged commits".
---
If your project is set up for it, you ca
Github user oknet closed the pull request at:
https://github.com/apache/trafficserver/pull/373
---
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
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/478
TS-4203: Multiple PD init/alloc in PollCont & setupPollDescriptor
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficse
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/477
TS-4202: missing define _HTTP_CONNECTION_COUNT_H_
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patch-6
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/476
TS-4201: call cont->handleEvent with OneWayTunnel object
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patc
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/400#issuecomment-172295039
my baseOS is Debian 7.8ï¼but the openssl upgraded to 1.0.2d with testing
---
If your project is set up for it, you can reply to this email and have your
reply
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/400#issuecomment-171212463
@bryancall the traffic.out log for current ATS release:
[Dec 28 21:01:12.742] Server {0x2b052fe4b700} DEBUG: (ssl) create ticket
for a new session
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/416
TS-4125: wrong non-active conditions check
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver patch-4
Alternatively you
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/400#issuecomment-169551690
please checking log message in the JIRA Issue TS-4104 for the different
between 0 and 1.
according the logs, handshake restart if return 0 or handshake
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/400
TS-4104: return 1 on a new ticket created
ref:
https://www.openssl.org/docs/manmaster/ssl/SSL_CTX_set_tlsext_ticket_key_cb.html
You can merge this pull request into a Git repository by
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/374#discussion_r48522460
--- Diff: iocore/net/SSLNetVConnection.cc ---
@@ -1095,6 +1105,12 @@ SSLNetVConnection::sslServerHandShakeEvent(int &
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/374#discussion_r48413673
--- Diff: iocore/net/SSLNetVConnection.cc ---
@@ -1400,10 +1416,9 @@ SSLNetVConnection::select_next_protocol(SSL *ssl,
const unsigned char **out, uns
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/373#issuecomment-167039829
Thanks! you are right. I got a core file that show me the write_vio is NULL
in gdb. Can you revert this pull request then create a new patch to set
write_vio in
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/374
TS-4075: add a state check for sslHandshakeHookState
Add a state check for sslHandshakeHookState after PreAcceptHookState
checking in sslServerHandShakeEvent().
and modify the codes in
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/373
TS-4055: InactivityCop callback SM with INACTIVE_TIMEOUT before
NET_EVENT_OPEN
for explain:
on a heavy load system, HttpSM create a connection to OS.
First, a TCP-SYN packet
Github user oknet commented on a diff in the pull request:
https://github.com/apache/trafficserver/pull/366#discussion_r46964787
--- Diff: iocore/net/UnixNetAccept.cc ---
@@ -112,9 +106,17 @@ net_accept(NetAccept *na, void *ep, bool blockable)
count = res
Github user oknet commented on the pull request:
https://github.com/apache/trafficserver/pull/366#issuecomment-162803942
Answersï¼
1. to implement an one shot accept (like a ftp pasv mode)
2. accept_fn = net_accept, accept_fn is called from NetAccept
3. the alloc_cache
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/366
remove alloc_cache from NetAccept
remove alloc_cache from NetAccept & net_accept() to avoid mem leak on
~NetAccept().
add a condition check for vc==NULL follow allocate_vc().
You
GitHub user oknet opened a pull request:
https://github.com/apache/trafficserver/pull/191
doc rst file patches
add missing file on ck/doc
rst format
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/oknet/trafficserver master
70 matches
Mail list logo
