Clodo: Suddenly the machine turned off - how to understand why?

Question

Debian

Today at 9.40 Moscow time the machine suddenly turned off (Debian, DC Oversan-Mercury). Searched the logs - nothing is clear.

Access to SSH from two IP addresses, one of which was active only at that moment (ubuntu is not Windows).

I thought about Claude herself - like xen theirs decided to turn something off. But tech support responded stupidly:

Hello
You should check the server logs. If nothing is indicated in the panel, it means that it could be turned off through the console.

> Do you have any logs from your side?
> Do you have any logs from your side? M.b. on your part, something fell or rebooted and did not take off?

No, unfortunately, we do not have such logs.

1. Is it true that the hoster does not keep any logs, how do virtual machines feel there?
2. Has anyone had a thread like this on the clod?
3. Where to look on debian or what to include in the future to understand what is wrong? The virtual machine itself works under xen - there is a xenmgm process inside.

Answer 1

Support answered you crookedly, we apologize for this. The person just did not look at all the logs. Your VPS caught a rather exotic migration bug. If you want, you can continue communication by mail [email protected]

Answer 2

1. I met the opposite, but I won’t say about everyone
2. And it didn’t happen like that! :) Check out the hub
3. Look at linode.com , www.digitalocean.com/ , joyent.com/

Answer 3

At one competing cloud service from St. Petersburg, by the way, reboots and shutdowns and their reason are displayed in the admin panel.

Answer 4

Can you post the number of the car? A very strange answer from the support
The machine is turned off either
- by the user through the API / applications / panel
- Cloud Core for non-payment / before scheduled work
, and this is all logged, or from the inside, but turning it off from the inside is followed by turning it on by Cloud Core.

Answer 5

What kind of strange phrase I looked for in the logs is not clear.
Specifically, provide the logs for the shutdown time + bash history.

Answer 6

The physical server crashes - there is no live migration / and if it is, it did not work. As a result, your vps is restarted on another physical server. Have you thought about this? (if there is an opportunity to check - at least for an exception). These are the clouds, and even what!

Answer 7

Today at 10:45 at the same Klodo, 2 VPS went down for an hour.