OVH Community, votre nouvel espace communautaire.

Multiple servers started to reboot periodically


ramil
08/12/2014, 20h44
Hello,

We're experiencing serious problems with servers stability and we've been advised to post on this forum by Lithuanian support team. At least 5 servers are affected.

The problem we encounter is episodical hard reboots of the servers during network activity (especially HTTP, but not only). SLA monitoring is disabled and this is not kernel panic (since we disabled reboot on kernel panic). Usually (but not always), this happens when the outgoing traffic is >200Mbit/s. Reboots may occur from several times a week to several times a day.

We have conducted several tests, opened different tickets, but the technicians have not figured out any problems on each server. We have tried different server models and different datacenters (SBG, RBX). Servers keep rebooting with standard OVH kernels.

This is not related to power supply, because if we try to route traffic through another server, only the server which responds clients is affected, even if there is no CPU load and it just routes traffic. Example:
Clients <-> XXX (router) <-> YYY (web server) === XXX reboots
Clients <-> YYY (router + web server) === YYY reboots

We suspect that it may be connected to the recent changes to anti-hack system (http://status.ovh.co.uk/?do=details&id=8120). But we do not receive any notifications on that (or any other) matter.

Please help us figure it out, because the situation is critical and we cannot work in such conditions. Actually, we have little time to do hardware diagnosis - it will not lead to any results. Can you confirm that reboots are initiated from OVH and list the situations that can lead to it?

Merci beacoup!

P.S. For support: ticket #1959468