Memory leak ?

[Hitman]

Hi,

I have a SPLAT Manager R60 HF03.

He is up for 28 days and the active real memory is at 1.02 Gig (always increasing ). Is this a bug (memory leak) or everyone has this huge amount of memory utilisation over days.

This manager only receive log from a cluster HA (2 firewall) and another stand alone firewall. Smart view monitor and the tracker are always running on a gui client.

This manager has 4 gig of ram, for now everything is fine, but what is going to happen if the memory reach 4 gig ??? in X number of days.....

Does anyone have the same problem ?

Thanks in advance

[melipla]

I believe it depends on the number of connections you have as to how much memory is used. Check /var/log/messages for any memory leak errors.

[Hitman]

I checked the file /var/log/messages and no error about memory leak or other
The only message that semms to be not normal is:

modprobe: modprobe: Can't locate module n

I did a top and the process fw is at 267 M the highest use of memory
fwm is at 74M

[Hitman]

I checked the file /var/log/messages and no error about memory leak or other.
The only message that semms to be not normal is:

modprobe: modprobe: Can't locate module n

I did a top and the process fw is at 267 M the highest use of memory
fwm is at 74M
cplmd is at 26M

Here is a copy of the memory utilisation of the top result


Mem: 4091352k av, 1070636k used, 3020716k free,

[markto]

I have R60A with HUGE memory leak. After failover, I reach 2GB within 2days. It then fails over, and is up to 2Gb again. Then fails over ........

Checkpoint says upgrade to R61. My top stats highest user is ahttpd. (R60A) so I assume the security server code is going nuts. I beleive this all started after a smart defense update sometime in late June. Anyone else using 60A?

[chillyjim]

I've heard rumors of memory leaks in R60A but it wasn't/isn't widely deployed.

AFAIK HFA03 is is fine. Please note that the underlying UNIX will use real memory for caching so it will often look like a memory leak from top.

BTW HFA4 has been released to the CSP's should be released to the public in about 2 weeks

[kva.kva]

Originally Posted by chillyjim

I've heard rumors of memory leaks in R60A but it wasn't/isn't widely deployed.

AFAIK HFA03 is is fine. Please note that the underlying UNIX will use real memory for caching so it will often look like a memory leak from top.

BTW HFA4 has been released to the CSP's should be released to the public in about 2 weeks

Is it possible to use R60 HFA (03, 04) with R60A?

[chillyjim]

Originally Posted by kva.kva

Is it possible to use R60 HFA (03, 04) with R60A?

No it is not

[cpgoof]

We have this with P-1 R65_HFA02 (incl. ENF_HF_HA02_249) on a 4BG SPLAT machine as well.....

[cciesec2006]

Originally Posted by cpgoof

We have this with P-1 R65_HFA03 (incl. ENF_HF_HA02_249) on a 4BG SPLAT machine as well.....

Checkpoint has released hfa_03 for NGx R65 already?

[cpgoof]

Originally Posted by cciesec2006

Checkpoint has released hfa_03 for NGx R65 already?

excuse me, just a typo....

[varera]

Originally Posted by markto

I have R60A with HUGE memory leak. After failover, I reach 2GB within 2days. It then fails over, and is up to 2Gb again. Then fails over ........

Checkpoint says upgrade to R61. My top stats highest user is ahttpd. (R60A) so I assume the security server code is going nuts. I beleive this all started after a smart defense update sometime in late June. Anyone else using 60A?

R60A is a minor release, and it won't be getting new HFAs

Upgrade to R62/R65.

Valeri Loukine
CCMA-0019