High Memory Usage + Random Crashes

[smiffy1989]

HI,

Just recently rebuilt my server switching from Unraid to WHS 2011. Using 0.7.1 and noticed that the sabnzbd.exe process is using upto 2GB of memory.

I'm also experiencing frequent but random crashes.

A few log snippets:

Code: Select all

2012-07-17 22:03:29,717::INFO::[__init__:826] Traceback: 

Traceback (most recent call last):
  File "sabnzbd\__init__.pyo", line 815, in save_data
MemoryError

Code: Select all

2012-07-17 22:02:53,198::INFO::[__init__:826] Traceback: 

Traceback (most recent call last):
  File "sabnzbd\__init__.pyo", line 815, in save_data
MemoryError

System is:
Core i3 with 4GB of RAM
WHS 2011
SABNZBD 0.7.1 using plush skin
Cache is set at the default 200M

Any ideas? Can provide any further information required

[shypike]

The root cause is a bug in Python's runtime, which should only occur
after a large amount of sizeable jobs.
The odd thing is that almost nobody reports this, so there may very well be
additional circumstances needed to trigger the bug.
One less-than-satisfactory remedy is to schedule a restart in SABnzbd every 24 hours or so.
We're looking at at an alternative implementation in a later release.
BTW: the large memory usage is new to me.

[smiffy1989]

My server has been taking a hammering, it has downloaded a couple of TB's over the last few days.

Sab has not crashed since I restarted it this morning.

An interesting thing is though, I have not accessed the webui (or used qouch) and the memory usage has stayed under 400MB.

I use IIS to reverse proxy which I should have probably mentioned previously.

At work atm but I will try do some testing when I get home to see if I can figure out what causes the high memory usage.