> Hi, > > At 17.38 17/05/2005, Holton, Euan wrote: > > >Hi fellow Squid users > > > >I've been using Squid NT v2.5 STABLE 7 successfully for some > months now, > >but I've noticed a few odd entries in cache.log, similar to the > >following: > > > >nntlm-atultmh-[a6u8t7h6[]4:6 5s2e]n: dsienngd i'nNgA 'LNoAg > oLno gfoani > >lfuariel:u ruen:k nuonwknn ouwsne ru snearm en aomre bor baad dp > >apsasswsowrodr.d'. 't ot os qsuqiudi > >D > > > >On first look , it appears to be random garbage, but on > looking closer > >it is apparent that it is two instances of the text "ntlm-auth[6876]: > >sending 'NA Logon failure: unknown user name or bad > password.' to squid" > >(with the [####] varying according to which ntlm-auth helper > is writing > >the log message) intermingled with each other, suggesting to > me that two > >different helpers are sending the same message to the log file at the > >same time (perhaps directly?). > > > >I've not had a chance to upgrade to the latest STABLE > version to see if > >this behaviour continues there, but in the meantime is there a simple > >change I could make to my squid.conf to avoid this problem in future? > > Never seen something similar before. It seems some problem on the > communication channel between squid.exe process and the > helpers. What was > the squid uptime ? > Do you rotate logs ? > > A restart of SquidNT service should stop the problem. > > Let me to know if this happens again. Squid was last restarted around 8am on 26 April, and generally isn't under a great deal of load (From cachemgr.cgi: UP Time: 1903558.654 seconds; CPU Time: 4763.953 seconds). The server running it also runs a couple of other services, none of which exercise the disk system particularly, and the box has 2 CPUs. All of the services the box runs are stable (ie, they've run for several weeks without need for tweaks or attention). This issue itself appears to be cosmetic, as Squid hasn't given us any problems. I currently rotate logs weekly, recently changed from daily (a couple of weeks ago), with logfile_rotate set to 14. The service barely pages, 90619 page faults with physical I/O. I may update to the latest binaries (2.5 STABLE 9) you offer on your download page. Are there any "gotchas" that should be considered before doing so? Thank you for the response! Regards Euan
