[jira] [Commented] (TS-1405) apply time-wheel scheduler about event system
[ https://issues.apache.org/jira/browse/TS-1405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13627875#comment-13627875 ] Leif Hedstrom commented on TS-1405: --- I think the max being down is an artifact of less pressure on the box (since it not can only do about 60% of the traffic it used to). I ran a few more tests, the second one tries to reduce the pressure on the box to verify that the max response time is due to the system being on its knees: With this patch, and 500 connections (there's not noticeable difference, other than mean time is 30% worse): {code} 6378965 fetches on 580129 conns, 498 max parallel, 6.378960E+08 bytes, in 60 seconds 100 mean bytes/fetch 106315.6 fetches/sec, 1.063156E+07 bytes/sec msecs/connect: 0.245 mean, 8.846 max, 0.042 min msecs/first-response: 3.791 mean, 207.045 max, 0.079 min {code} Current master with 300 connections, but at a lower QPS (so less pressure): {code} 8850329 fetches on 8 conns, 300 max parallel, 8.850330E+08 bytes, in 60 seconds 100 mean bytes/fetch 147505.5 fetches/sec, 1.475055E+07 bytes/sec msecs/connect: 0.191 mean, 2.037 max, 0.043 min msecs/first-response: 0.678 mean, 77.340 max, 0.085 min {code} So even though this second test on master is doing significantly more QPS (almost 50% more), it still has much better response response times across the board. By reducing the throughput in this last test, such that the system resources aren't at their limits, the response times improve. I think that's why with the patch, you see slightly better response times on Max, but it's really not indicative of the patch improving anything. It's because with the patch, ATS simply can't put the system under pressure. This is pretty much the same problem I posted about early on here. As far as I can tell, it's gotten noticeably worse since the first patch sets :). apply time-wheel scheduler about event system -- Key: TS-1405 URL: https://issues.apache.org/jira/browse/TS-1405 Project: Traffic Server Issue Type: Improvement Components: Core Affects Versions: 3.2.0 Reporter: Bin Chen Assignee: Bin Chen Fix For: 3.3.2 Attachments: linux_time_wheel.patch, linux_time_wheel_v10jp.patch, linux_time_wheel_v11jp.patch, linux_time_wheel_v2.patch, linux_time_wheel_v3.patch, linux_time_wheel_v4.patch, linux_time_wheel_v5.patch, linux_time_wheel_v6.patch, linux_time_wheel_v7.patch, linux_time_wheel_v8.patch, linux_time_wheel_v9jp.patch when have more and more event in event system scheduler, it's worse. This is the reason why we use inactivecop to handler keepalive. the new scheduler is time-wheel. It's have better time complexity(O(1)) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (TS-1405) apply time-wheel scheduler about event system
[ https://issues.apache.org/jira/browse/TS-1405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13627875#comment-13627875 ] Leif Hedstrom edited comment on TS-1405 at 4/10/13 3:05 PM: I think the max being down is an artifact of less pressure on the box (since it now can only do about 60% of the traffic it used to). I ran a few more tests, the second one tries to reduce the pressure on the box to verify that the max response time is due to the system being on its knees: With this patch, and 500 connections (there's not noticeable difference, other than mean time is 30% worse): {code} 6378965 fetches on 580129 conns, 498 max parallel, 6.378960E+08 bytes, in 60 seconds 100 mean bytes/fetch 106315.6 fetches/sec, 1.063156E+07 bytes/sec msecs/connect: 0.245 mean, 8.846 max, 0.042 min msecs/first-response: 3.791 mean, 207.045 max, 0.079 min {code} Current master with 300 connections, but at a lower QPS (so less pressure): {code} 8850329 fetches on 8 conns, 300 max parallel, 8.850330E+08 bytes, in 60 seconds 100 mean bytes/fetch 147505.5 fetches/sec, 1.475055E+07 bytes/sec msecs/connect: 0.191 mean, 2.037 max, 0.043 min msecs/first-response: 0.678 mean, 77.340 max, 0.085 min {code} So even though this second test on master is doing significantly more QPS (almost 50% more), it still has much better response response times across the board. By reducing the throughput in this last test, such that the system resources aren't at their limits, the response times improve. I think that's why with the patch, you see slightly better response times on Max, but it's really not indicative of the patch improving anything. It's because with the patch, ATS simply can't put the system under pressure. This is pretty much the same problem I posted about early on here. As far as I can tell, it's gotten noticeably worse since the first patch sets :). was (Author: zwoop): I think the max being down is an artifact of less pressure on the box (since it not can only do about 60% of the traffic it used to). I ran a few more tests, the second one tries to reduce the pressure on the box to verify that the max response time is due to the system being on its knees: With this patch, and 500 connections (there's not noticeable difference, other than mean time is 30% worse): {code} 6378965 fetches on 580129 conns, 498 max parallel, 6.378960E+08 bytes, in 60 seconds 100 mean bytes/fetch 106315.6 fetches/sec, 1.063156E+07 bytes/sec msecs/connect: 0.245 mean, 8.846 max, 0.042 min msecs/first-response: 3.791 mean, 207.045 max, 0.079 min {code} Current master with 300 connections, but at a lower QPS (so less pressure): {code} 8850329 fetches on 8 conns, 300 max parallel, 8.850330E+08 bytes, in 60 seconds 100 mean bytes/fetch 147505.5 fetches/sec, 1.475055E+07 bytes/sec msecs/connect: 0.191 mean, 2.037 max, 0.043 min msecs/first-response: 0.678 mean, 77.340 max, 0.085 min {code} So even though this second test on master is doing significantly more QPS (almost 50% more), it still has much better response response times across the board. By reducing the throughput in this last test, such that the system resources aren't at their limits, the response times improve. I think that's why with the patch, you see slightly better response times on Max, but it's really not indicative of the patch improving anything. It's because with the patch, ATS simply can't put the system under pressure. This is pretty much the same problem I posted about early on here. As far as I can tell, it's gotten noticeably worse since the first patch sets :). apply time-wheel scheduler about event system -- Key: TS-1405 URL: https://issues.apache.org/jira/browse/TS-1405 Project: Traffic Server Issue Type: Improvement Components: Core Affects Versions: 3.2.0 Reporter: Bin Chen Assignee: Bin Chen Fix For: 3.3.2 Attachments: linux_time_wheel.patch, linux_time_wheel_v10jp.patch, linux_time_wheel_v11jp.patch, linux_time_wheel_v2.patch, linux_time_wheel_v3.patch, linux_time_wheel_v4.patch, linux_time_wheel_v5.patch, linux_time_wheel_v6.patch, linux_time_wheel_v7.patch, linux_time_wheel_v8.patch, linux_time_wheel_v9jp.patch when have more and more event in event system scheduler, it's worse. This is the reason why we use inactivecop to handler keepalive. the new scheduler is time-wheel. It's have better time complexity(O(1)) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (TS-1405) apply time-wheel scheduler about event system
[ https://issues.apache.org/jira/browse/TS-1405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13627875#comment-13627875 ] Leif Hedstrom edited comment on TS-1405 at 4/10/13 3:06 PM: I think the max being down is an artifact of less pressure on the box (since it now can only do about 60% of the traffic it used to). I ran a few more tests, the second one tries to reduce the pressure on the box to verify that the max response time is due to the system being on its knees: With this patch, and 500 connections (there's not noticeable difference, other than mean time is 30% worse): {code} 6378965 fetches on 580129 conns, 498 max parallel, 6.378960E+08 bytes, in 60 seconds 100 mean bytes/fetch 106315.6 fetches/sec, 1.063156E+07 bytes/sec msecs/connect: 0.245 mean, 8.846 max, 0.042 min msecs/first-response: 3.791 mean, 207.045 max, 0.079 min {code} Current master with 300 connections, but at a lower QPS (so less pressure): {code} 8850329 fetches on 8 conns, 300 max parallel, 8.850330E+08 bytes, in 60 seconds 100 mean bytes/fetch 147505.5 fetches/sec, 1.475055E+07 bytes/sec msecs/connect: 0.191 mean, 2.037 max, 0.043 min msecs/first-response: 0.678 mean, 77.340 max, 0.085 min {code} So even though this second test on master is doing significantly more QPS (almost 50% more), it still has much better response response times across the board. By reducing the throughput in this last test, such that the system resources aren't at their limits (and probably less rescheduling on lock contention), the response times improve. I think that's why with the patch, you see slightly better response times on Max, but it's really not indicative of the patch improving anything. It's because with the patch, ATS simply can't put the system under pressure. This is pretty much the same problem I posted about early on here. As far as I can tell, it's gotten noticeably worse since the first patch sets :). was (Author: zwoop): I think the max being down is an artifact of less pressure on the box (since it now can only do about 60% of the traffic it used to). I ran a few more tests, the second one tries to reduce the pressure on the box to verify that the max response time is due to the system being on its knees: With this patch, and 500 connections (there's not noticeable difference, other than mean time is 30% worse): {code} 6378965 fetches on 580129 conns, 498 max parallel, 6.378960E+08 bytes, in 60 seconds 100 mean bytes/fetch 106315.6 fetches/sec, 1.063156E+07 bytes/sec msecs/connect: 0.245 mean, 8.846 max, 0.042 min msecs/first-response: 3.791 mean, 207.045 max, 0.079 min {code} Current master with 300 connections, but at a lower QPS (so less pressure): {code} 8850329 fetches on 8 conns, 300 max parallel, 8.850330E+08 bytes, in 60 seconds 100 mean bytes/fetch 147505.5 fetches/sec, 1.475055E+07 bytes/sec msecs/connect: 0.191 mean, 2.037 max, 0.043 min msecs/first-response: 0.678 mean, 77.340 max, 0.085 min {code} So even though this second test on master is doing significantly more QPS (almost 50% more), it still has much better response response times across the board. By reducing the throughput in this last test, such that the system resources aren't at their limits, the response times improve. I think that's why with the patch, you see slightly better response times on Max, but it's really not indicative of the patch improving anything. It's because with the patch, ATS simply can't put the system under pressure. This is pretty much the same problem I posted about early on here. As far as I can tell, it's gotten noticeably worse since the first patch sets :). apply time-wheel scheduler about event system -- Key: TS-1405 URL: https://issues.apache.org/jira/browse/TS-1405 Project: Traffic Server Issue Type: Improvement Components: Core Affects Versions: 3.2.0 Reporter: Bin Chen Assignee: Bin Chen Fix For: 3.3.2 Attachments: linux_time_wheel.patch, linux_time_wheel_v10jp.patch, linux_time_wheel_v11jp.patch, linux_time_wheel_v2.patch, linux_time_wheel_v3.patch, linux_time_wheel_v4.patch, linux_time_wheel_v5.patch, linux_time_wheel_v6.patch, linux_time_wheel_v7.patch, linux_time_wheel_v8.patch, linux_time_wheel_v9jp.patch when have more and more event in event system scheduler, it's worse. This is the reason why we use inactivecop to handler keepalive. the new scheduler is time-wheel. It's have better time complexity(O(1)) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (TS-1719) Lua Plugin breaks build on Linux (Ubuntu 12.10/amd64)
[ https://issues.apache.org/jira/browse/TS-1719?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] James Peach updated TS-1719: Description: {noformat} Making all in lua make[3]: Entering directory `/home/igalic/src/asf/trafficserver/plugins/experimental/lua' /bin/bash ../../../libtool --tag=CXX --mode=compile g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c -o lua_la-state.lo `test -f 'state.cc' || echo './'`state.cc libtool: compile: g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c state.cc -fPIC -DPIC -o .libs/lua_la-state.o state.cc: In function 'instanceid_t LuaPluginRegister(unsigned int, const char**)': state.cc:169:21: error: comparison between signed and unsigned integer expressions [-Werror=sign-compare] state.cc: In member function 'bool LuaThreadState::init(LuaPluginInstance*)': state.cc:251:7: error: 'strerror' was not declared in this scope cc1plus: all warnings being treated as errors make[3]: *** [lua_la-state.lo] Error 1 {noformat} was: {noformat} Making all in lua make[3]: Entering directory `/home/igalic/src/asf/trafficserver/plugins/experimental/lua' /bin/bash ../../../libtool --tag=CXX --mode=compile g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c -o lua_la-state.lo `test -f 'state.cc' || echo './'`state.cc libtool: compile: g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c state.cc -fPIC -DPIC -o .libs/lua_la-state.o state.cc: In function 'instanceid_t LuaPluginRegister(unsigned int, const char**)': state.cc:169:21: error: comparison between signed and unsigned integer expressions [-Werror=sign-compare] state.cc: In member function 'bool LuaThreadState::init(LuaPluginInstance*)': state.cc:251:7: error: 'strerror' was not declared in this scope cc1plus: all warnings being treated as errors make[3]: *** [lua_la-state.lo] Error 1 {format} Lua Plugin breaks build on Linux (Ubuntu 12.10/amd64) - Key: TS-1719 URL: https://issues.apache.org/jira/browse/TS-1719 Project: Traffic Server Issue Type: Bug Components: Build Reporter: Igor Galić Assignee: James Peach Fix For: 3.3.2 {noformat} Making all in lua make[3]: Entering directory `/home/igalic/src/asf/trafficserver/plugins/experimental/lua' /bin/bash ../../../libtool --tag=CXX --mode=compile g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c -o lua_la-state.lo `test -f 'state.cc' || echo './'`state.cc libtool: compile: g++ -DHAVE_CONFIG_H -I. -I../../../lib/ts -I/usr//include/luajit-2.0 -I../../../proxy/api -I../../../proxy/api -D_LARGEFILE64_SOURCE=1 -D_COMPILE64BIT_SOURCE=1 -D_GNU_SOURCE -D_REENTRANT -Dlinux -I/usr/include/tcl8.5 -std=c++11 -g -pipe -Wall -Werror -O3 -feliminate-unused-debug-symbols -fno-strict-aliasing -Wno-invalid-offsetof -MT lua_la-state.lo -MD -MP -MF .deps/lua_la-state.Tpo -c state.cc -fPIC -DPIC -o .libs/lua_la-state.o state.cc: In function 'instanceid_t LuaPluginRegister(unsigned int, const char**)': state.cc:169:21: error: comparison between signed and unsigned integer expressions [-Werror=sign-compare] state.cc: In member
[jira] [Updated] (TS-608) Is HttpSessionManager::purge_keepalives() too aggressive?
[ https://issues.apache.org/jira/browse/TS-608?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Leif Hedstrom updated TS-608: - Fix Version/s: (was: 3.3.2) 3.3.3 Is HttpSessionManager::purge_keepalives() too aggressive? -- Key: TS-608 URL: https://issues.apache.org/jira/browse/TS-608 Project: Traffic Server Issue Type: Bug Components: HTTP Reporter: Leif Hedstrom Fix For: 3.3.3 Attachments: TS-608.patch It seems that if we trigger the max server connections, we call this purge function in the session manager, which will close all currently open keep-alive connections. This seems very aggressive, why not limit it to say only removing 10% of each bucket or some such? Also, how does this work together with per-origin limits? Ideally, if the per-origin limits are in place, we would only purge sessions that are for the IP we wish to connect to ? -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (TS-1053) get combo_handler compiled
[ https://issues.apache.org/jira/browse/TS-1053?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13628487#comment-13628487 ] Leif Hedstrom commented on TS-1053: --- So, I have some patches for this. but, I'm also thinking we should move this plugin in under the esi plugin source tree. Such that it builds both esi.so and combo_handler.so. They both share code, and it makes sense to combine them I think. get combo_handler compiled -- Key: TS-1053 URL: https://issues.apache.org/jira/browse/TS-1053 Project: Traffic Server Issue Type: Task Components: Plugins Reporter: Conan Wang Assignee: Leif Hedstrom Fix For: 3.3.4 Attachments: combo_handler.diff, fetcher.diff, Makefile combo_handler require ESI's code. Before make ESI work as a lib, you can try it this way: make esi/lib and esi/fetcher the subdir of combo_handler and use the makefile. {noformat} combo_handler |combo_handler.cc |fetcher |lib |LICENSE |Makefile |README {noformat} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (TS-1053) get combo_handler compiled
[ https://issues.apache.org/jira/browse/TS-1053?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Leif Hedstrom updated TS-1053: -- Fix Version/s: (was: 3.3.4) 3.3.2 get combo_handler compiled -- Key: TS-1053 URL: https://issues.apache.org/jira/browse/TS-1053 Project: Traffic Server Issue Type: Task Components: Plugins Reporter: Conan Wang Assignee: Leif Hedstrom Fix For: 3.3.2 Attachments: combo_handler.diff, fetcher.diff, Makefile combo_handler require ESI's code. Before make ESI work as a lib, you can try it this way: make esi/lib and esi/fetcher the subdir of combo_handler and use the makefile. {noformat} combo_handler |combo_handler.cc |fetcher |lib |LICENSE |Makefile |README {noformat} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (TS-1779) Crash using SNI and ssl_ca_name
[ https://issues.apache.org/jira/browse/TS-1779?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13628491#comment-13628491 ] Rodney commented on TS-1779: It happens everytime, so very easy to reproduce. Let me know how you want to receive it. Crash using SNI and ssl_ca_name --- Key: TS-1779 URL: https://issues.apache.org/jira/browse/TS-1779 Project: Traffic Server Issue Type: Bug Components: SSL Reporter: Rodney Assignee: James Peach Fix For: 3.3.2 When I add 'ssl_ca_name' to include a chain cert CA the traffic server fails to start with a core dump. It seems to be okay if I just have one entry in 'ssl_multicert.config' file but as soon as I use SNI the traffic server will not start with a core dump. This witnessed on 3.2.0 and currently 3.2.4 with Debian Squeeze. Example entries: ssl_cert_name=my1.crt ssl_key_name=my1.key ssl_ca_name=my1CA.crt ssl_cert_name=my2.crt ssl_key_name=my2.key ssl_ca_name=my2CA.crt #Default dest_ip=* ssl_cert_name=my1.crt ssl_key_name=my1.key ssl_ca_name=my1CA.crt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (TS-1779) Crash using SNI and ssl_ca_name
[ https://issues.apache.org/jira/browse/TS-1779?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13628602#comment-13628602 ] James Peach commented on TS-1779: - There should be a stack trace in the log file, probably traffic.out. That would actually be more useful than the core file, since I probably won't have the matching toolchain to symbolicate it. If you have a throwaway set of certificates you can give me, that would be great too :) Crash using SNI and ssl_ca_name --- Key: TS-1779 URL: https://issues.apache.org/jira/browse/TS-1779 Project: Traffic Server Issue Type: Bug Components: SSL Reporter: Rodney Assignee: James Peach Fix For: 3.3.2 When I add 'ssl_ca_name' to include a chain cert CA the traffic server fails to start with a core dump. It seems to be okay if I just have one entry in 'ssl_multicert.config' file but as soon as I use SNI the traffic server will not start with a core dump. This witnessed on 3.2.0 and currently 3.2.4 with Debian Squeeze. Example entries: ssl_cert_name=my1.crt ssl_key_name=my1.key ssl_ca_name=my1CA.crt ssl_cert_name=my2.crt ssl_key_name=my2.key ssl_ca_name=my2CA.crt #Default dest_ip=* ssl_cert_name=my1.crt ssl_key_name=my1.key ssl_ca_name=my1CA.crt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (TS-1405) apply time-wheel scheduler about event system
[ https://issues.apache.org/jira/browse/TS-1405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13628638#comment-13628638 ] Bin Chen commented on TS-1405: -- http_load -parallel 100 -seconds 60 -keep_alive 100 /tmp/URL all /tmp/URL is hit or miss? how about hit ratio? apply time-wheel scheduler about event system -- Key: TS-1405 URL: https://issues.apache.org/jira/browse/TS-1405 Project: Traffic Server Issue Type: Improvement Components: Core Affects Versions: 3.2.0 Reporter: Bin Chen Assignee: Bin Chen Fix For: 3.3.2 Attachments: linux_time_wheel.patch, linux_time_wheel_v10jp.patch, linux_time_wheel_v11jp.patch, linux_time_wheel_v2.patch, linux_time_wheel_v3.patch, linux_time_wheel_v4.patch, linux_time_wheel_v5.patch, linux_time_wheel_v6.patch, linux_time_wheel_v7.patch, linux_time_wheel_v8.patch, linux_time_wheel_v9jp.patch when have more and more event in event system scheduler, it's worse. This is the reason why we use inactivecop to handler keepalive. the new scheduler is time-wheel. It's have better time complexity(O(1)) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (TS-1405) apply time-wheel scheduler about event system
[ https://issues.apache.org/jira/browse/TS-1405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13628664#comment-13628664 ] Leif Hedstrom commented on TS-1405: --- 100% cache hit ratio. Not that I run 3x of those http_load, for a total of 300 connections. apply time-wheel scheduler about event system -- Key: TS-1405 URL: https://issues.apache.org/jira/browse/TS-1405 Project: Traffic Server Issue Type: Improvement Components: Core Affects Versions: 3.2.0 Reporter: Bin Chen Assignee: Bin Chen Fix For: 3.3.2 Attachments: linux_time_wheel.patch, linux_time_wheel_v10jp.patch, linux_time_wheel_v11jp.patch, linux_time_wheel_v2.patch, linux_time_wheel_v3.patch, linux_time_wheel_v4.patch, linux_time_wheel_v5.patch, linux_time_wheel_v6.patch, linux_time_wheel_v7.patch, linux_time_wheel_v8.patch, linux_time_wheel_v9jp.patch when have more and more event in event system scheduler, it's worse. This is the reason why we use inactivecop to handler keepalive. the new scheduler is time-wheel. It's have better time complexity(O(1)) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (TS-1106) redirect map generates multiple Via: header entries.
[ https://issues.apache.org/jira/browse/TS-1106?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Leif Hedstrom updated TS-1106: -- Fix Version/s: (was: 3.3.3) 3.5.0 redirect map generates multiple Via: header entries. Key: TS-1106 URL: https://issues.apache.org/jira/browse/TS-1106 Project: Traffic Server Issue Type: Bug Components: HTTP Affects Versions: 3.1.2 Reporter: Leif Hedstrom Assignee: Leif Hedstrom Fix For: 3.5.0 It seems using the redirect rule in remap.config, ends up duplicating the entry in the Via: header, e.g. {code} Via: http/1.1 kramer.ogre.com (ApacheTrafficServer/3.1.3-unstable [u c s f p eS:tNc i p s ]), http/1.1 kramer.ogre.com (ApacheTrafficServer/3.1.3-unstable [u c s f p eS:tNc i p s ]) {code} I'm not sure why this happens, if it's duplicating it, or if it's going through the SM twice. I know i'm not proxying twice through the box. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira