Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
On Wed, 11 May 2005, Sally Huang wrote: Thanks a lot. I send to a wrong email address last time. Where is the lastest epoll patch that solved the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Is it in squid src tagged with epoll-2.5 in squid CVS? Or, is it in the patch from http://devel.squid-cache.org/cgi-bin/diff2/epoll-2_5.patch?s2_5? They are the same. Regards Henrik
RE: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
On Tue, 10 May 2005, Sally Huang wrote: Guess it is the same problem with Pawe Staszewski. Steven said he hasfound the solution and updated CVS. May I know where I can get the updated patch? Shall I download the whole squid src tagged with epoll-2.5 in squid CVS? Or, may I just download the newest squid2.5STABLE10 version, and applied the patch from http://devel.squid-cache.org/cgi-bin/diff2/epoll-2_5.patch?s2_5? Either method is fine. Regards Henrik
Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
Thanks for your clarify. The author doesn't reply me. Could you pls do me a favor and check with the author whether he has updated the latest epoll patch so that everyone won't face this epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Regards, sally On 5/10/05, Henrik Nordstrom [EMAIL PROTECTED] wrote: On Tue, 10 May 2005, Sally Huang wrote: Do you mean the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 bug isalready fixed on both epoll-2_5 patch and epoll-2.5 branch in squidCVS? I am the wrong person to answer that question, but as you I remember the branch author mentioning that there has been fixes in that area. I answerd to the question how you could get the latest version of the epoll branch. Regards Henrik
RE: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
I did manage to trace the cause of the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 messages. The latest epoll patch fixes this problem. There are 2 messages that the new patch will cause to appear with the default debug options (ALL,1): storeClientCopy3: url - clearing ENTRY_DEFER_READ This is caused when an object has been deferred, and for some reason the memory has not been freed, even though all clients have seen all in-memory parts of the url. I can't see why this would be the case, and I'm sure it's not caused by the epoll code, but the epoll code needs to catch this condition. WARNING defer handler for fd=fd(url) does not call commDeferFD() - backing off manually. This is caused when an object has data ready to be read, and the defer handler reports that the read should be deferred, but does not tell the epoll code to back off. This debug would probably be very noisy if used with delay pools, but other than that it indicates a non-optimised defer handler. These two debug statements do not produce a large number of entries in the debug logs (181 and 52 messages respectively out of 3204 messages total in one of our proxy server's cache.log out of a total of ~4.2 million requests). The proxy server does handle these cases appropriately, but they indicate sub-optimal performance. As they represent a tiny fraction of the total number of requests, I have not spent any time working out the exact cause of these problems. Regards Steven -Original Message- From: Sally Huang [mailto:[EMAIL PROTECTED] Sent: Wednesday, May 11, 2005 9:34 AM To: Henrik Nordstrom Cc: squid-users@squid-cache.org Subject: Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9 Thanks for your clarify. The author doesn't reply me. Could you pls do me a favor and check with the author whether he has updated the latest epoll patch so that everyone won't face this epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Regards, sally On 5/10/05, Henrik Nordstrom [EMAIL PROTECTED] wrote: On Tue, 10 May 2005, Sally Huang wrote: Do you mean the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 bug isalready fixed on both epoll-2_5 patch and epoll-2.5 branch in squidCVS? I am the wrong person to answer that question, but as you I remember the branch author mentioning that there has been fixes in that area. I answerd to the question how you could get the latest version of the epoll branch. Regards Henrik -- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005 -- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005
Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
Thanks a lot. I send to a wrong email address last time. Where is the lastest epoll patch that solved the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Is it in squid src tagged with epoll-2.5 in squid CVS? Or, is it in the patch from http://devel.squid-cache.org/cgi-bin/diff2/epoll-2_5.patch?s2_5? sally On 5/11/05, Steven Wilton [EMAIL PROTECTED] wrote: I did manage to trace the cause of the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 messages. The latest epoll patch fixes this problem. There are 2 messages that the new patch will cause to appear with the default debug options (ALL,1): storeClientCopy3: url - clearing ENTRY_DEFER_READ This is caused when an object has been deferred, and for some reason the memory has not been freed, even though all clients have seen all in-memory parts of the url. I can't see why this would be the case, and I'm sure it's not caused by the epoll code, but the epoll code needs to catch this condition. WARNING defer handler for fd=fd(url) does not call commDeferFD() - backing off manually. This is caused when an object has data ready to be read, and the defer handler reports that the read should be deferred, but does not tell the epoll code to back off. This debug would probably be very noisy if used with delay pools, but other than that it indicates a non-optimised defer handler. These two debug statements do not produce a large number of entries in the debug logs (181 and 52 messages respectively out of 3204 messages total in one of our proxy server's cache.log out of a total of ~4.2 million requests). The proxy server does handle these cases appropriately, but they indicate sub-optimal performance. As they represent a tiny fraction of the total number of requests, I have not spent any time working out the exact cause of these problems. Regards Steven -Original Message- From: Sally Huang [mailto:[EMAIL PROTECTED] Sent: Wednesday, May 11, 2005 9:34 AM To: Henrik Nordstrom Cc: squid-users@squid-cache.org Subject: Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9 Thanks for your clarify. The author doesn't reply me. Could you pls do me a favor and check with the author whether he has updated the latest epoll patch so that everyone won't face this epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Regards, sally On 5/10/05, Henrik Nordstrom [EMAIL PROTECTED] wrote: On Tue, 10 May 2005, Sally Huang wrote: Do you mean the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 bug isalready fixed on both epoll-2_5 patch and epoll-2.5 branch in squidCVS? I am the wrong person to answer that question, but as you I remember the branch author mentioning that there has been fixes in that area. I answerd to the question how you could get the latest version of the epoll branch. Regards Henrik -- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005 -- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005
RE: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
The updated patch is available in both locations that you have mentioned. CVS does get updated quicker than the web site (the web site sync's with CVS periodically). Steven -Original Message- From: Sally Huang [mailto:[EMAIL PROTECTED] Sent: Wednesday, May 11, 2005 12:03 PM Thanks a lot. I send to a wrong email address last time. Where is the lastest epoll patch that solved the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Is it in squid src tagged with epoll-2.5 in squid CVS? Or, is it in the patch from http://devel.squid-cache.org/cgi-bin/diff2/epoll-2_5.patch?s2_5? sally On 5/11/05, Steven Wilton [EMAIL PROTECTED] wrote: I did manage to trace the cause of the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 messages. The latest epoll patch fixes this problem. There are 2 messages that the new patch will cause to appear with the default debug options (ALL,1): storeClientCopy3: url - clearing ENTRY_DEFER_READ This is caused when an object has been deferred, and for some reason the memory has not been freed, even though all clients have seen all in-memory parts of the url. I can't see why this would be the case, and I'm sure it's not caused by the epoll code, but the epoll code needs to catch this condition. WARNING defer handler for fd=fd(url) does not call commDeferFD() - backing off manually. This is caused when an object has data ready to be read, and the defer handler reports that the read should be deferred, but does not tell the epoll code to back off. This debug would probably be very noisy if used with delay pools, but other than that it indicates a non-optimised defer handler. These two debug statements do not produce a large number of entries in the debug logs (181 and 52 messages respectively out of 3204 messages total in one of our proxy server's cache.log out of a total of ~4.2 million requests). The proxy server does handle these cases appropriately, but they indicate sub-optimal performance. As they represent a tiny fraction of the total number of requests, I have not spent any time working out the exact cause of these problems. Regards Steven -Original Message- From: Sally Huang [mailto:[EMAIL PROTECTED] Sent: Wednesday, May 11, 2005 9:34 AM To: Henrik Nordstrom Cc: squid-users@squid-cache.org Subject: Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9 Thanks for your clarify. The author doesn't reply me. Could you pls do me a favor and check with the author whether he has updated the latest epoll patch so that everyone won't face this epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 problem? Regards, sally On 5/10/05, Henrik Nordstrom [EMAIL PROTECTED] wrote: On Tue, 10 May 2005, Sally Huang wrote: Do you mean the epoll_ctl(EPOLL_CTL_DEL): failed on fd=197 bug isalready fixed on both epoll-2_5 patch and epoll-2.5 branch in squidCVS? I am the wrong person to answer that question, but as you I remember the branch author mentioning that there has been fixes in that area. I answerd to the question how you could get the latest version of the epoll branch. Regards Henrik -- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005 -- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005 -- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005 -- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.8 - Release Date: 5/10/2005
Re: Odp: RE: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
On Sat, 16 Apr 2005, [Windows-1250] Pawe Staszewski wrote: Hello edit the /usr/local/share/autoconf/acconfig.h file and add the folowing line : #undef HAVE_EPOLL No, this should be done in the Squid sources, same filename. Regards Henrik
Odp: RE: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
Hello edit the /usr/local/share/autoconf/acconfig.h file and add the folowing line : #undef HAVE_EPOLL Best Regards Pawe Staszewski ART-COM +48327522333 +480609183038 George Hong [EMAIL PROTECTED] 04/16/05 12:33 pm Hi there, When I run the bootstrap.sh it gives me the following error mesg: /usr/local/bin/autoheader: Symbol `HAVE_EPOLL' is not covered by /usr/local/share/autoconf/acconfig.h ./acconfig.h autoheader failed I have the kernel 2.6.7.11(it should include epoll support) installed on the Redhat. What did I miss? Thanks for your suggestion. BR George -Original Message- From: Henrik Nordstrom [mailto:[EMAIL PROTECTED] Sent: Tuesday, April 12, 2005 11:08 PM To: Pawe?Staszewski Cc: squid-users@squid-cache.org Subject: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9 On Tue, 12 Apr 2005, [Windows-1250] Pawe Staszewski wrote: and then in squid dir i do : patch -p1 epoll-2_5.patch Patch was added cleanly but: ./configure --help | grep epoll show nothing :( You need to run the bootstrap.sh script after patching with developer patches. This in turn requires you to have the proper versions of autoconf and automake installed (it warns you if it cannot find the proper versions) Regards Henrik
Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
On Wed, 13 Apr 2005, [Windows-1250] Pawe Staszewski wrote: When i apply epool-2_5.patch In my syslog i have this messages: Apr 13 09:42:25 w3cache1 squid[24895]: commSetSelect: epoll_ctl(EPOLL_CTL_DEL): failed on fd=197: (2) No such file or directory Apr 13 09:42:25 w3cache1 squid[24895]: commSetSelect: epoll_ctl(EPOLL_CTL_DEL): failed on fd=234: (2) No such file or directory Looks like the epoll patch is not 100% perfect yet. Regards Henrik
Re: Odp: Re: [squid-users] how to apply epoll-2_5 patch to squid2.5-stable9
Ok :) Some comparisions: ## Squid-2.5Stable9-20050330 with epoll 12:09:58 [EMAIL PROTECTED] /home/paol/squid-2.5.STABLE9-20050330]$ mpstat -P ALL 1 Linux 2.6.12-rc2 (w3cache1) 04/13/05 12:12:36 CPU %user %nice %system %iowait%irq %soft %idle intr/s 12:12:37 all1.500.000.503.500.000.50 94.00 2484.16 12:12:37 00.990.000.993.960.000.00 94.06 991.09 12:12:37 11.980.000.992.970.001.98 91.09 1493.07 12:12:37 CPU %user %nice %system %iowait%irq %soft %idle intr/s 12:12:38 all1.000.001.002.490.000.50 95.02 2620.20 12:12:38 00.000.000.003.030.000.00 96.97 1011.11 12:12:38 12.020.002.021.010.000.00 95.96 1609.09 12:12:38 CPU %user %nice %system %iowait%irq %soft %idle intr/s 12:12:39 all1.010.001.011.010.001.01 95.96 2809.09 12:12:39 00.000.000.001.010.000.00 98.99 1011.11 12:12:39 13.030.001.011.011.013.03 92.93 1797.98 12:12:39 CPU %user %nice %system %iowait%irq %soft %idle intr/s 12:12:40 all1.490.000.995.450.501.49 90.10 2834.65 12:12:40 00.000.000.998.910.000.00 90.10 991.09 12:12:40 11.980.001.982.970.002.97 90.10 1843.56 12:12:40 CPU %user %nice %system %iowait%irq %soft %idle intr/s 12:12:41 all2.000.002.00 23.000.001.50 71.50 2839.60 12:12:41 01.980.000.99 29.700.000.00 67.33 991.09 12:12:41 12.970.001.98 15.840.001.98 74.26 1848.51 [12:12:42 [EMAIL PROTECTED] /home/paol/squid-2.5.STABLE9-20050330]$ [12:12:42 [EMAIL PROTECTED] /home/paol/squid-2.5.STABLE9-20050330]$ iostat 1 -x Linux 2.6.12-rc2 (w3cache1) 04/13/05 avg-cpu: %user %nice%sys %iowait %idle 4.840.00 14.995.33 74.84 Device:rrqm/s wrqm/s r/s w/s rsec/s wsec/srkB/swkB/s avgrq-sz avgqu-sz await svctm %util sda 0.05 17.29 4.09 11.33 46.48 239.8323.24 119.9118.56 0.41 26.64 3.42 5.27 sdb 0.05 13.22 3.99 7.02 45.44 172.9722.7286.4819.84 0.24 21.62 3.61 3.98 sdc 0.04 13.39 3.32 6.59 38.68 170.4519.3485.2321.10 0.20 20.22 3.53 3.50 sdd 0.05 12.95 4.03 6.68 46.13 167.9023.0683.9519.98 0.26 24.05 3.65 3.91 avg-cpu: %user %nice%sys %iowait %idle 1.000.001.001.00 97.00 Device:rrqm/s wrqm/s r/s w/s rsec/s wsec/srkB/swkB/s avgrq-sz avgqu-sz await svctm %util sda 0.00 31.68 0.00 30.690.00 511.88 0.00 255.9416.68 0.062.10 0.61 1.88 sdb 0.00 26.73 0.00 3.960.00 279.21 0.00 139.6070.50 0.000.75 0.75 0.30 sdc 0.00 15.84 0.99 2.977.92 187.13 3.9693.5649.25 0.049.00 9.00 3.56 sdd 0.00 123.76 2.97 12.87 31.68 1165.3515.84 582.6775.56 0.20 12.75 3.94 6.24 avg-cpu: %user %nice%sys %iowait %idle 1.490.002.490.50 95.52 Device:rrqm/s wrqm/s r/s w/s rsec/s wsec/srkB/swkB/s avgrq-sz avgqu-sz await svctm %util sda 0.00 0.00 2.02 18.18 24.24 141.4112.1270.71 8.20 0.136.35 4.15 8.38 sdb 0.00 0.00 4.04 7.07 48.48 42.4224.2421.21 8.18 0.043.91 3.27 3.64 sdc 0.00 0.00 6.06 1.01 64.65 64.6532.3232.3218.29 0.034.00 4.00 2.83 sdd 0.00 0.00 3.03 3.03 24.24 17.1712.12 8.59 6.83 0.011.50 1.17 0.71 avg-cpu: %user %nice%sys %iowait %idle 1.000.001.491.49 96.02 Device:rrqm/s wrqm/s r/s w/s rsec/s wsec/srkB/swkB/s avgrq-sz avgqu-sz await svctm %util sda 0.00 0.00 0.00 0.000.000.00 0.00 0.00 0.00 0.000.00 0.00 0.00 sdb 0.00 0.00 2.00 0.00 24.000.0012.00 0.0012.00 0.013.00 3.00 0.60 sdc 0.00 0.00 7.00 3.00 88.00 17.0044.00 8.5010.50 0.033.30 3.10 3.10 sdd 0.00 0.00 6.00 0.00 64.000.0032.00 0.0010.67 0.046.00 6.00 3.60 [12:13:38 [EMAIL PROTECTED] /home/paol/squid-2.5.STABLE9-20050330]$ Connection information for squid: Number of clients accessing cache: 0 Number of HTTP