Re: Memory leak with ldirectord 1.88 on fedora core 1

[James Bromberger <jbromberger@xxxxxxxxxxx>]

I saw the same thing, and Arthur Bergman here wrote a simple cut down 
of the test used in ldirectord to get to the error, somewhere in 
Net::SSLeay. See:
        http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=254671


Its the SSL test you have. Try either plicating the SSL test by only 
doing it once every so often (checktype=60 to do it once per minute), 
or just do a connect test (checktype=connect).
HTH,

  James


On 4 Aug 2004, at 01:32, Jonathan Trott wrote:

> We have had ldirectord 1.88 running on our Fedora Core 1 cluster for a 
> few
> months now with occasional crashes of the active server. I have tracked
> this down to a memory leak, probably in perl, that is causing 
> ldirectord
> to take up more and more memory on the machine until it runs out and
> starts killing processes.
> For example, after 8 days uptime ldirectord was consuming 487MB of RAM
> until I rebooted it this morning.
> Does anyone have any idea where I should start looking to patch perl to
> fix this problem?
> Perl is 5.8.3 that comes with Fedora Core 1. I have already checked 
> CPAN
> for updates to Net::SSLeay as that has come up before with leakages and
> the version that is installed (1.26) is much newer than the version 
> that
> had a memory leak (1.20 or 1.21). For completeness, this is my
> ldirectord.cf:
>
> checktimeout=6
> checkinterval=2
> autoreload=yes
> #logfile="/var/log/ldirectord.log"
> logfile="local0"
> quiescent=no
>
> virtual=192.168.100.1:80
>         real=192.168.11.11:80 masq
>         real=192.168.11.12:80 masq
>         real=192.168.11.13:80 masq
>         real=192.168.11.14:80 masq
>         fallback=127.0.0.1:80
>         service=http
>         request="/serverstate/"
>         receive="server up"
>         scheduler=rr
>         #persistent=5
>         #netmask=255.255.255.255
>         protocol=tcp
>
> virtual=192.168.0.1:80
>         real=192.168.11.11:80 masq
>         real=192.168.11.12:80 masq
>         real=192.168.11.13:80 masq
>         real=192.168.11.14:80 masq
>         fallback=127.0.0.1:80
>         service=http
>         request="/serverstate/"
>         receive="server up"
>         scheduler=rr
>         #persistent=5
>         #netmask=255.255.255.255
>         protocol=tcp
>
> virtual=192.168.100.1:443
>         real=192.168.11.11:443 masq
>         real=192.168.11.12:443 masq
>         real=192.168.11.13:443 masq
>         real=192.168.11.14:443 masq
>         fallback=127.0.0.1:443
>         service=https
>         request="/serverstate/"
>         receive="server up"
>         scheduler=rr
>         #persistent=5
>         #netmask=255.255.255.255
>         protocol=tcp
>
> virtual=192.168.0.1:443
>         real=192.168.11.11:443 masq
>         real=192.168.11.12:443 masq
>         real=192.168.11.13:443 masq
>         real=192.168.11.14:443 masq
>         fallback=127.0.0.1:443
>         service=https
>         request="/serverstate/"
>         receive="server up"
>         scheduler=rr
>         #persistent=5
>         #netmask=255.255.255.255
>         protocol=tcp
>
> Thanks,
> JT
> _______________________________________________
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://www.in-addr.de/mailman/listinfo/lvs-users