
je at instadia
Nov 2, 2004, 8:50 AM
Post #1 of 2
(2191 views)
Permalink
|
|
Ldirectord check_https problems
|
|
Hi Ldirectord specialists, I have big problems monitor https-services, causing all https-services to be down. >From debugging ldirectord, I have the following. -------------- snippet of debug-print start sslcat 11324: sending 70 bytes... write_all VM at entry=vm_unknown written so far 70:70 bytes (VM=vm_unknown) waiting for reply... got 602:0 bytes (VM=vm_unknown). got 0:602 bytes (VM=vm_unknown). Got 602 bytes. DEBUG2: Result: HTTP/1.1 200 OK DEBUG2: Status: 16777215 DEBUG2: Disabled server=10.5.1.21 -------------- stop It seems that the child-process are doing what it's supposed to. The webpage is found, retrieved and the receive-string matches. It then returns 0. So far everything is OK. However, in the master-process it doesn't get this status - instead the value is -1. Shifting this >> 8 gives us 16777215 - which is not equal to 0 - and therefore the server is disabled. I'm not a perl programmer, but I think my problem has somethig to do with the forking. If I comment out the line: local $SIG{'CHLD'} = "IGNORE"; in "sub check_https". It seems to work. Then status 0 is fetched, and the https service made available. What function does this line have? Is it Ok to remove it? or are there any other workarounds I should try. I'm running Debian (sarge) with Perl 5.8.4-2.3, and a 2.6.8.1-kernel. Regards, Johan -- Johan Elmerfjord, CTO Instadia A/S Islands Brygge 43 DK-2300 Copenhagen Tel. +45 36 98 89 55 Fax. +45 36 98 89 51 http://www.instadia.net _______________________________________________ Linux-HA mailing list Linux-HA[at]lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha
|