[squid-users] Squid + kerberos, all childrens are busy

Дмитрий Рукавцов 2005now at mail.ru
Mon Jul 13 17:54:13 UTC 2015






    >>>>> Hello, i have a problem here :) System - freebsd 10.1, squid 3.5.5 + kerberos (MIT), 50 users total.
    >>>>>
    >>>>> Without any auth my squid works fine, system is not loaded. When i enable Kerberos auth internet slowly goes down and crushing after a while, at logs i see:
    >>>>>
    >>>>> 2015/07/09 11:47:14 kid1| WARNING: All 60/60 negotiateauthenticator processes are busy.
    >>>>> 2015/07/09 11:47:14 kid1| WARNING: 72 pending requests queued
    >>>>>
    >>>>
    >>>> So 50 users / 60 helpers ... how many requests per second? and how
    >>>> fast/slow is the helper responding?
    >> Could you clarify how I can get value of requests per second and respond?
    >
    >The cachemgr "info" report. From the cachemgr.cgi tool, or "squidclient
    >mgr:info" command line, or
    >http://$visible_hostname:3128/squid-internal-mgr/info
    >
    > Or calculated from a quick count of the access.log lines over a few mins.

    ~600 lines per minute,



    >> Debugs show like 3-4 message per second like:
    >>
    >> negotiate_kerberos_pac.cc(376): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Got PAC data of lengh 624
    >> negotiate_kerberos_pac.cc(180): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Found 5 rids
    >> negotiate_kerberos_pac.cc(188): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: Info: Got rid: 513
    >> negotiate_kerberos_pac.cc(188): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: Info: Got rid: 3692
    >> negotiate_kerberos_pac.cc(188): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: Info: Got rid: 4268
    >> negotiate_kerberos_pac.cc(188): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: Info: Got rid: 4622
    >> negotiate_kerberos_pac.cc(188): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: Info: Got rid: 4623
    >> negotiate_kerberos_pac.cc(256): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Got DomainLogonId S-1-5-21-1708537768-1580818891-2146958067
    >> negotiate_kerberos_pac.cc(278): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Found 1 ExtraSIDs
    >> negotiate_kerberos_pac.cc(327): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Got ExtraSid S-1-5-21-1708537768-1580818891-2146958067-4107
    >> negotiate_kerberos_pac.cc(456): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: INFO: Read 620 of 624 bytes
    >> negotiate_kerberos_auth.cc(778): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: DEBUG: Groups group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/AQIAAA== group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/bA4AAA== group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/rBAAAA== group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/DhIAAA== group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/DxIAAA== group=AQUAAAAAAAUVAAAAqDfWZcthOV7z+vd/CxAAAA==
    >> negotiate_kerberos_auth.cc(783): pid=1456 :2015/07/09 13:26:48| negotiate_kerberos_auth: DEBUG: AF oYGyMIGvoAMKAQChCwYJKoZIgvcSAQICooGaBIGXYIGUBgkqhkiG9xIBAgICAG+BhDCBgaADAgEFoQMCAQ+idTBzoAMCAReibARqY4fSYtg+X4HhiH8dFmWxdn3wxtoKKZzEfUjLYibMoy0XAAWgkSYVXgC7gxO7cgCkOofEqZQhi/GKa4NZqn2dQqOJU/3y4zkPqBP9Ialh//BL5ov03L5BqjgthrbYbrcxJTo57EJIdO8O1g== avialex
    >>
    >> And errors like:
    >> 2015/07/09 13:28:03 kid1| ERROR: Negotiate Authentication validating user. Result: {result=BH, notes={message: received type 1 NTLM token; }}
    >> All my friends get the same error, but their squid is working fine.
    >>
    >> Don't see anything else
    >>
    >
    >Aha. So your users browsers are sending NTLM auth instead of Kerberos.
    >That is at least one part of the problem. NTLM handshake can take whole
    >seconds and places a lot of extra load on the helpers. To resolve these
    >the users software needs fixing to use Kerberos properly when Negotiate

    >is offered.
    >
    >The other part is figuring out what amount of helpers is needed to meet
    >the load requirements. With NTLM it is usually several hundred.
    >


When i'm using proxy alone, squid stars 33 childrens, don't recive any NTLM errors, but internet start to lag. So the problem not in the NTLM software. I tryed to start 300 children for my 60 users, but still have huge lags, even when half was free.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squid-cache.org/pipermail/squid-users/attachments/20150713/22ed30ad/attachment.html>


More information about the squid-users mailing list