Approaching the limit on PV entries
-
Any help from anybody please… :'( :'( :'(
-
Have you tried a web search on this topic?
The first hit of my favorite search engine give a solution: https://forums.freebsd.org/threads/17786/
-
Yes I've been through this site. And from their advice the default value of vm.pmap.shpgperproc=200 should be increased but I already have an value of 500 I am not sure if should I still need to increase this. Dont really know what drawback we get if I modified this, cant risk it, we dont have a backup plan if this router went down.
-
looks like you on ancient unsupported version as that variable (and limit) is gone from FreeBSD for about 3-4 years now.
It doesnt exist in FreeBSD 9 and newer.
-
Now pfSense crashes just by now and we dont have the log message anymore on the console "Approaching the limit on PV entries, consider increasing either the vm.pmap.shpgperproc or the vm.pmap.pv_entry_max tunable". Was this the reason why our pfsense crashed?
We are currently running a 2.2.6-RELEASE (i386) FreeBSD 10.1-RELEASE-p25
-
We are currently running a 2.2.6-RELEASE (i386) FreeBSD 10.1-RELEASE-p25
unless someone has encountered this in the past, you won't get many replies. Almost none of the frequent posters / developers run an out-dated version.
the only post i've found that is related is from august 2015 & remained without resolution.
https://forum.pfsense.org/index.php?topic=98799 -
That's bad news then. Well I guess my only hope now will be upgrading my firewall to a latest pfSense version
-
I still use version 2.2.6 in multiple VMs, running solid 200+ days no reboot, and they have been rock solid since install. One thing I noticed is that you are running on 32 bit. If you want to stay on 2.2.6 I would recommend trying 64-bit if it is supported by your system, that is what we use.
-
I still use version 2.2.6 in multiple VMs, running solid 200+ days no reboot, and they have been rock solid since install. One thing I noticed is that you are running on 32 bit. If you want to stay on 2.2.6 I would recommend trying 64-bit if it is supported by your system, that is what we use.
Our planned resolution to this is to install a new pfsense 2.3.2 version hopefully this logs wont occur again.
Thanks everyone for your attention. -
Just ran across this on one of my firewall appliances.
Approaching the limit on PV entries, consider increasing either the vm.pmap.shpgperproc or the vm.pmap.pv_entries tunable.What exactly is a "PV" entry? No one said.
This firewall has been running for 840 days and just started dropping some phonecalls for the site this week, so i am just going to reboot it and see if this error goes away. But it would be interesting to know what exactly this error message indicates. What can cause this behaviour?
From this article here: https://supportportal.juniper.net/s/article/EX2300-C-printing-logs-Approaching-the-limit-on-PV-entries?language=en_US
Seems to have something to do with virtual memory.
On the kernel, each virtual address which is mapped to a given dynamically-allocated physical memory page uses up one pv_entry in the kernel, which contains information about that specific virtual translation. The kernel imposes a max limit on the number of pv entries that can be allocated on the system.
This message means the system is running low on pv entries in the kernel,
will post back if rebooting fixes it.
device is an SG-3100 running 21.05.2-RELEASE