Project

General

Profile

Actions

Bug #2791

closed

OOM errors on low end devices

Added by Shivani Bhardwaj almost 6 years ago. Updated almost 6 years ago.

Status:
Closed
Priority:
High
Target version:
Affected Versions:
Effort:
Difficulty:
Label:

Description

As per a recent discussion on the mailing list, suricata-update takes up too much memory to run and causes OOM errors on low end devices like RPi.

Actions #1

Updated by Shivani Bhardwaj almost 6 years ago

  • Status changed from New to Assigned
Actions #2

Updated by Victor Julien almost 6 years ago

Perhaps there are multiple issues. The 'test' step where Suricata itself validates the new rules spins up a new Suricata next to a running one. This will lead to duplicate mem use for Suricata itself. This would not be suricata-update's fault of course.

Actions #3

Updated by Jason Ish almost 6 years ago

  • Status changed from Assigned to Closed
  • Target version changed from TBD to 1.0.4

Marking closed for now. I think we've addressed all the low hanging fruit items we can without a redesign of the internals. Feedback from the user shows he is able to update the rules, without these changes, but has to use the --no-test parameter.

https://github.com/OISF/suricata-update/commit/5cea9cf4f29fca00ede5c0882f2f9356415f3aba

Actions #4

Updated by Konstantin Klinger almost 6 years ago

I think not only the rule testing is the problem here. Our suricata-update instance that runs on a seperate rulehost-server with 2GB RAM gets killed while running suricata-update, because lack of available memory.

This happens during the following line:
7/3/2019 -- 10:01:35 - <Info> -- Backing up current rules.

I think the diff function between the current rule file and the new one consumes to much memory and leads to the killing of the process.

Actions #5

Updated by Victor Julien almost 6 years ago

One of the things that can make a difference is making sure that you use python3 instead of python2.

Actions

Also available in: Atom PDF