[squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-05 Thread Victor Sudakov
Rafael Akchurin wrote: > I believe I do (but you made me doubt:) > Well, I have tried negotiate_kerberos_auth with Firefox (Windows) and they don't work together. I am attaching a packet dump which boils down basically to the following: 1. proxy.sibptus.transneft.ru:3131 is configured in Firefox

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-05 Thread Victor Sudakov
And before I forget and before somebody asks. In Firefox: network.negotiate-auth.allow-proxies=true network.negotiate-auth.gsslib="" network.negotiate-auth.using-native-gsslib=true Victor Sudakov wrote: > Rafael Akchurin wrote: > > I believe I do (but you made me doubt:) > > > > Well, I have tr

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-05 Thread Victor Sudakov
Victor Sudakov wrote: > Rafael Akchurin wrote: > > I believe I do (but you made me doubt:) > > > > Well, I have tried negotiate_kerberos_auth with Firefox (Windows) I have tried the same with MSIE 8 (Windows). It's obviously trying to do NTLM instead of Kerberos (see below). How do I enable Ke

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Victor Sudakov
Victor Sudakov wrote: > > Well, I have tried negotiate_kerberos_auth with Firefox (Windows) > > I have tried the same with MSIE 8 (Windows). After some adjustment to domain group policies, the Windows host is at last requesting and successfully receiving the ticket for the proxy service. Wiresh

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 And my Kerberos server setup seems valid: $ setenv KRB5_KTNAME /usr/local/etc/squid/squid.keytab $ setenv KRB5_CONFIG /usr/local/etc/squid/krb5.conf $ kdestroy $ kinit -t $KRB5_KTNAME HTTP/proxy.sibptus.transneft.ru $ klist Credentials cache: FILE:/t

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Victor Sudakov
And my Kerberos server setup seems valid: $ setenv KRB5_KTNAME /usr/local/etc/squid/squid.keytab $ setenv KRB5_CONFIG /usr/local/etc/squid/krb5.conf $ kdestroy $ kinit -t $KRB5_KTNAME HTTP/proxy.sibptus.transneft.ru $ klist Credentials cache: FILE:/tmp/krb5cc_Ld5uU9 Principal: HTTP/proxy.

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Colleagues, I am posting below the contents of an HTTP request (especially the "Proxy-Authorization:" header the browser is sending) to which squid's negotiate_kerberos_auth replies: "ERROR: Negotiate Authentication validating user. Result: {result=

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Markus Moeller
Hi Victor, In the helpers/negotiate_auth/kerberos directory is a script test_negotiate_auth.sh to test authentication outside of squid. Change dir to your binary directory and do the following ( please adapt to your environment): export KRB5_KTNAME=squid-win.keytab kinit m...@win2003r2.hom

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Markus Moeller wrote: > > In the helpers/negotiate_auth/kerberos directory is a script > test_negotiate_auth.sh to test authentication outside of squid. Markus, I could find the said script neither in the source nor in the binary package. Howeve

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-07 Thread Eliezer Croitoru
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 10/08/2014 06:29 AM, Victor Sudakov wrote: > Markus, > > I could find the said script neither in the source nor in the > binary package. However I think I can guess what could be inside. > Could you look below if that makes sense? Or you can just

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-08 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Eliezer Croitoru wrote: > > > > I could find the said script neither in the source nor in the > > binary package. However I think I can guess what could be inside. > > Could you look below if that makes sense? > > Or you can just look at the source c

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-08 Thread Markus Moeller
Hi Victor, I only found the following explanation: This error will happen if you didn't write the key into the keytab file, or the permission setting of keytab file reject the read access, or the key file is not the one you should access (for example, you want /opt/somedir/conf/krb5.conf,

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-08 Thread Markus Moeller
Hi Victor, I only found the following explanation: This error will happen if you didn't write the key into the keytab file, or the permission setting of keytab file reject the read access, or the key file is not the one you should access (for example, you want /opt/somedir/conf/krb5.conf, but

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-08 Thread Victor Sudakov
Markus Moeller wrote: > >I only found the following explanation: > > This error will happen if you didn't write the key into the keytab file, or > the permission setting of keytab file reject the read access, or the key > file is not the one you should access (for example, you want > /opt/som

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-10 Thread Victor Sudakov
Colleagues, What if the service principal's name in squid's keytab does not coincide with the host's primary FQDN (AKA `hostname`)? E.g. the squid's keytab contains keys for HTTP/proxy.my.domain while the server's actual FQDN is fw.my.domain? Should it cause the obscure error I have stumbled up

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-10 Thread Markus Moeller
I think it could. Can you try the option -s GSS_C_NO_NAME ? Markus "Victor Sudakov" wrote in message news:20141010113630.ga39...@admin.sibptus.tomsk.ru... Colleagues, What if the service principal's name in squid's keytab does not coincide with the host's primary FQDN (AKA `hostname`)? E.g

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-10 Thread Victor Sudakov
Markus Moeller wrote: > > > What if the service principal's name in squid's keytab does not > > coincide with the host's primary FQDN (AKA `hostname`)? > > > > E.g. the squid's keytab contains keys for HTTP/proxy.my.domain while > > the server's actual FQDN is fw.my.domain? > > > > Should it cau

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-11 Thread Markus Moeller
Good to see it works now. As far as I recall the MIT message is clearer in this case. Regards Markus "Victor Sudakov" wrote in message news:20141011044626.gb49...@admin.sibptus.tomsk.ru... Markus Moeller wrote: > What if the service principal's name in squid's keytab does not > coincide w

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-11 Thread Markus Moeller
Hi Viktor, These sections of code do the selection in squid: char *service_name = (char *) "HTTP", *host_name = NULL; if (service_principal && strcasecmp(service_principal, "GSS_C_NO_NAME")) { service.value = service_principal; service.length = strlen((char *) service.value); } else {

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-11 Thread Victor Sudakov
Markus Moeller wrote: > Hi Viktor, > >These sections of code do the selection in squid: > > > char *service_name = (char *) "HTTP", *host_name = NULL; Thanks for posting this. BTW does it mean that the service name "HTTP" is hardcoded, and if I wanted to use a principal with a different ser

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-11 Thread Markus Moeller
HTTP is the standard service for HTTP authentication (web and proxy) Markus "Victor Sudakov" wrote in message news:20141011131747.ga56...@admin.sibptus.tomsk.ru... Markus Moeller wrote: Hi Viktor, These sections of code do the selection in squid: char *service_name = (char *) "HTTP",

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-11 Thread Markus Moeller
Also you can overwrite it with the -s option if you really need to. Markus "Victor Sudakov" wrote in message news:20141011131747.ga56...@admin.sibptus.tomsk.ru... Markus Moeller wrote: Hi Viktor, These sections of code do the selection in squid: char *service_name = (char *) "HTTP", *

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-13 Thread Victor Sudakov
Now as I run the test authenticator, what's the "INFO: continuation needed" message from the squid Kerberos helper? How do I interperet it? Is it success or fail? See output below: $ bin/test_negotiate_auth.sh /home/sudakov/squid.keytab: Vno Type Principal 1 des-cbc-crc

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-14 Thread Markus Moeller
Hi Victor, That just means that the server requires more information from the client. This could happen if mutual authentication is required or the dataset is too large and had to be split. If you run it in squid the client would send new data until the server says the exchange is complete (

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-16 Thread Victor Sudakov
This question is neither exactly squid-related nor Heimdal-related, but maybe someone guru could shed some light. I configure MSIE to use the proxy server "proxy.sibptus.transneft.ru". On starting MSIE, some Windows hosts request a ticket for the principal HTTP/proxy.sibptus.transneft.ru" and rec

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-16 Thread Markus Moeller
Hi Victor, That sounds a bit strange. Can you capture with wireshark the traffic on port 88 on the system which has squiduser in the cache ( best after a clear the cache with kerbtray first) when accessing squid and send it to me as cap file ? Markus "Victor Sudakov" wrote in message n

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-16 Thread Victor Sudakov
Markus Moeller wrote: > > That sounds a bit strange. Can you capture with wireshark the traffic on > port 88 on the system which has squiduser in the cache ( best after a clear > the cache with kerbtray first) when accessing squid and send it to me as cap > file ? I am attaching a traffic dum

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-16 Thread Eugene M. Zheganin
Hi. On 17.10.2014 11:02, Victor Sudakov wrote: > > I am attaching a traffic dump. > > Please look at Frame No. 36, where a ticket is requested for > "HTTP/proxy.sibptus.transneft.ru", and then at Frame No. 39, where > the ticket is granted, but for the wrong principal name. > The thing is, valid e

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-17 Thread Victor Sudakov
Markus Moeller wrote: > > > > Now as I run the test authenticator, what's the > > "INFO: continuation needed" > > message from the squid Kerberos helper? How do I interperet it? Is it > > success or fail? > > That just means that the server requires more information from the client. > This cou

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-18 Thread Victor Sudakov
Eugene M. Zheganin wrote: > > > > I am attaching a traffic dump. > > > > Please look at Frame No. 36, where a ticket is requested for > > "HTTP/proxy.sibptus.transneft.ru", and then at Frame No. 39, where > > the ticket is granted, but for the wrong principal name. > > > The thing is, valid exchang

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-18 Thread Eugene M. Zheganin
Hi. On 18.10.2014 16:11, Victor Sudakov wrote: I thought as much. This error seems suspicious. But why does a second request not cause the same error? No idea. We have tried both ways (enabling all ciphers and enabling only arcfour-hmac-md5), but it made no difference. Currently we are using t

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-19 Thread Victor Sudakov
Eugene M. Zheganin wrote: > > On 18.10.2014 16:11, Victor Sudakov wrote: > > I thought as much. This error seems suspicious. But why does a second > > request not cause the same error? > No idea. Hopefully I can interest our Windows admin to enable Kerberos event logging per KB262177. But for th

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-19 Thread Eliezer Croitoru
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 10/19/2014 10:32 AM, Victor Sudakov wrote: > Hopefully I can interest our Windows admin to enable Kerberos > event logging per KB262177. > > But for the present I have found an ugly workaround. In squid's > keytab, I created another principal calle

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-19 Thread Eugene M. Zheganin
Hi. On 19.10.2014 13:32, Victor Sudakov wrote: > > Hopefully I can interest our Windows admin to enable Kerberos event > logging per KB262177. > > But for the present I have found an ugly workaround. In squid's keytab, I > created another principal called 'squiduser' with the same hex key and > kv

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-20 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Eugene M. Zheganin wrote: > > > > Hopefully I can interest our Windows admin to enable Kerberos event > > logging per KB262177. > > > > But for the present I have found an ugly workaround. In squid's keytab, I > > created another principal called 'squi

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-20 Thread Eugene M. Zheganin
Hi. On 20.10.2014 22:29, Victor Sudakov wrote: That's what we did. 1. Created an AD user called squiduser. 2. Extracted its keytab with something like ktpass -princ HTTP/proxy.sibptus.transneft...@sibptus.transneft.ru -mapuser squiduser +rndPass -out squid.keytab -ptype KRB5_NT_PRINCIPAL /t

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-20 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Eliezer Croitoru wrote: > > Hopefully I can interest our Windows admin to enable Kerberos > > event logging per KB262177. > > > > But for the present I have found an ugly workaround. In squid's > > keytab, I created another principal called 'squiduser

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-22 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Eliezer Croitoru wrote: > > And about the basic issues that you were having with performance, does > it help to run Kerberos instead of NTLM (it should...)? I have even moved squid to a new virtual machine (FreeBSD 9.3-RELEASE under VMWare, 1 GB RAM)

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-22 Thread Amos Jeffries
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 23/10/2014 5:53 p.m., Victor Sudakov wrote: > Eliezer Croitoru wrote: > >> And about the basic issues that you were having with performance, >> does it help to run Kerberos instead of NTLM (it should...)? > > I have even moved squid to a new virtu

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-22 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Amos Jeffries wrote: > > > >> And about the basic issues that you were having with performance, > >> does it help to run Kerberos instead of NTLM (it should...)? > > > > I have even moved squid to a new virtual machine (FreeBSD > > 9.3-RELEASE under

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-22 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Victor Sudakov wrote: > > > > > >> And about the basic issues that you were having with performance, > > >> does it help to run Kerberos instead of NTLM (it should...)? > > > > > > I have even moved squid to a new virtual machine (FreeBSD > > > 9.3-R

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Amos Jeffries
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 23/10/2014 7:27 p.m., Victor Sudakov wrote: > Victor Sudakov wrote: > And about the basic issues that you were having with > performance, does it help to run Kerberos instead of NTLM > (it should...)? I have even moved s

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Eliezer Croitoru wrote: > > I don't know what's happening with squid but this kind of CPU > > consumption is just not normal: > > > > Victor are you using workers by any chance? I doubt it. I have compiled it from the FreeBSD ports, there was no "w

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Amos Jeffries wrote: > > And about the basic issues that you were having with > > performance, does it help to run Kerberos instead of NTLM > > (it should...)? > > I have even moved squid to a new virtual machine (FreeBSD >

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Carlos Defoe
I had this kind of 100% CPU problem with auth helpers when upgrading to squid 3.4. I use negotiate_wrapper, kerberos, ntlm and basic auth. Then I had to fall back to 3.3 and it is production until now, with some troubles with broken clients, but with normal cpu usage most of the time. Can you try w

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Eugene M. Zheganin
Hi. On 23.10.2014 18:13, Carlos Defoe wrote: > I had this kind of 100% CPU problem with auth helpers when upgrading > to squid 3.4. I use negotiate_wrapper, kerberos, ntlm and basic auth. > Then I had to fall back to 3.3 and it is production until now, with > some troubles with broken clients, but

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-23 Thread Carlos Defoe
no, rhel 6 On Thu, Oct 23, 2014 at 9:51 AM, Eugene M. Zheganin wrote: > Hi. > > On 23.10.2014 18:13, Carlos Defoe wrote: >> I had this kind of 100% CPU problem with auth helpers when upgrading >> to squid 3.4. I use negotiate_wrapper, kerberos, ntlm and basic auth. >> Then I had to fall back to 3

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-10-24 Thread Victor Sudakov
Carlos Defoe wrote: > I had this kind of 100% CPU problem with auth helpers when upgrading > to squid 3.4. I use negotiate_wrapper, kerberos, ntlm and basic auth. > Then I had to fall back to 3.3 and it is production until now, with > some troubles with broken clients, but with normal cpu usage mos

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-11-03 Thread Victor Sudakov
Colleagues, I have created a howto in Russian about squid and Kerberos proxy authentication, addressing also the two problems I personally encountered while setting all the stuff up. If any Russian speakers here could review and comment it, I would be grateful. The text is at http://victor-suda

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-11-04 Thread Victor Sudakov
Now available at https://bitbucket.org/victor_sudakov/faq/src/tip/FAQ/squid_kerberos.txt Victor Sudakov wrote: > Colleagues, > > I have created a howto in Russian about squid and Kerberos proxy > authentication, addressing also the two problems I personally > encountered while setting all the st

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-11-04 Thread Amos Jeffries
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 4/11/2014 9:10 p.m., Victor Sudakov wrote: > Now available at > https://bitbucket.org/victor_sudakov/faq/src/tip/FAQ/squid_kerberos.txt > > Victor Sudakov wrote: >> Colleagues, >> >> I have created a howto in Russian about squid and Kerberos proxy

Re: [squid-users] Squid, Kerberos and FireFox (Was: Re: leaking memory in squid 3.4.8 and 3.4.7.)

2014-11-04 Thread Victor Sudakov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Amos Jeffries wrote: > On 4/11/2014 9:10 p.m., Victor Sudakov wrote: > > Now available at > > https://bitbucket.org/victor_sudakov/faq/src/tip/FAQ/squid_kerberos.txt [dd] > > If you are going to publish this please use either the official > domains