FreeNAS using too much ram/swap and stalling/freezing?

Status
Not open for further replies.

risho

Dabbler
Joined
May 21, 2016
Messages
18
My FreeNAS box seems to be randomly crashing when i use it. I have 32gb ram, a Intel(R) Xeon(R) CPU E3-1231 v3 @ 3.40GHz for the specs. I'm running a plex jail and a rutorrent jail. i have an ubuntu vm that i use with docker that does a log of non-heavy stuff in docker instances. plexpy, calibre-rdp, a secondary rutorrent instance, a shoko instance(it's an anime cataloguing thing.)

here's what my ipmi showed right after it stopped functioning.
[redacted]
i had to hard reset it using the ipmi controls.

now the vm won't start at all. I've rolled back the dataset to before the crash, I've created a new vm and used a clone of an old snapshot as the hard disk and all i get is this:


here's some screenshots of the graphs up to the point where it stopped functioning and after i booted it back up again.
[redacted]
it's unclear what it is that is using all the memory, but i wonder if it was the vm. i can't really test it as it exploded.

here's a graph of the memory/swap over a longer time period.
[redacted]

while i would really like to find the issue that was causing them memory to go out of control, i would really like to get this vm working again...

let me know if there are any logs or screenshots that i could post that would offer useful information.
 

Attachments

  • danger.PNG
    danger.PNG
    112.7 KB · Views: 284
  • dangercpu.PNG
    dangercpu.PNG
    55.7 KB · Views: 301
  • dangermem.PNG
    dangermem.PNG
    33.8 KB · Views: 270
  • dangermemlongtime.PNG
    dangermemlongtime.PNG
    144.9 KB · Views: 281
  • dangerzfs.PNG
    dangerzfs.PNG
    339.3 KB · Views: 284
  • vnc.PNG
    vnc.PNG
    14.6 KB · Views: 305
Last edited by a moderator:

deastick

Dabbler
Joined
Dec 27, 2013
Messages
26
I've gotten that same swap error twice since upgrading from 9.3 to 11.1-RELEASE. I run plex, samba, and some other plugins (sickbeard, etc). Not high demands. I have not figured out the cause yet on my system. Hopefully will figure it out to be able to submit a ticket.
 

danb35

Hall of Famer
Joined
Aug 16, 2011
Messages
15,504
Please don't post links to external image-hosting sites--just paste the images into your post, or otherwise upload them here.

A number of people have reported similar issues. Reverting to 11.0 is the only solution I've seen at this point.
 

risho

Dabbler
Joined
May 21, 2016
Messages
18
i'm more than happy to go back to 11.0, but is there a way for me to recover my vm?
 

deastick

Dabbler
Joined
Dec 27, 2013
Messages
26
There is a pretty nasty memory leak bug in the FreeBSD kernel version 11.1 uses.
This bug bit me on a FreeBSD VM I have running. It has only 1GB RAM so it ran out of memory very quickly.
Check : https://redmine.ixsystems.com/issues/27270
I am postponing upgrading from 11.0 to 11.1 until 11.1-U1 is available.

Paul
Yeah thanks. I read that last night. I'll keep an eye on things and reboot every day or two until that release comes out .

Sent from my SM-G920W8 using Tapatalk
 
Joined
Jan 4, 2018
Messages
2
I can confirm this is a major bug as well. That is what I was getting on here to ask about. Thank you for the information. I upgraded from 11.0-U1 to 11.1 and am experiencing this on my production box. The replication box, which is basically the exact same hardware is not seeing this issue, and has all the same services, drivers, hardware, etc running as well. That means it has something to do with either the SMB or NFS process IMHO eating all the memory.
Here is what I have been seeing. The low point around week 50 is when I upgraded to 11.1
 
Last edited by a moderator:

Ericloewe

Server Wrangler
Moderator
Joined
Feb 15, 2014
Messages
20,194
@GreatWhiteDan, linking to an image on your local network is not going to do us any good. Copy and paste it into the forum.
 
Status
Not open for further replies.
Top