K
K
Krio2012-05-06 01:29:13
Hard disks
Krio, 2012-05-06 01:29:13

Checking hard drive on VPS?

Good afternoon.

I keep a small test vps at hetzner (for testing and studying operating systems), on Friday (05/04/2012) logging in via SSH debian gave out that I connected in read-only mode. At first I thought that the place was over, but only 15% was occupied and there were a lot of errors during scanning that the file could not be read.
When connecting via a virtual console, I noticed a message that there are bad sectors on the screw, during reboot the axis generally fell off.

Due to this, the question arose, what tools are better to check the status of the disk?
I would like to check for cron and send the report to email.

Thank you.

PS who has VPS at hetzner - check :)

Answer the question

In order to leave comments, you need to log in

4 answer(s)
A
amarao, 2012-05-06
@amarao

(as a person who knows where this “read only” comes from)
In R/O, the file system falls down due to disk failures. This is usually the timeout for a write/read operation (for a virtual environment). This usually happens for one of two reasons:
* the initiator's iscsi timeout at the host
* its own timeout inside the operating system (about 2m)
What should I do?
a) check with the support whether it is possible to restart the machine. In some cases, this is highly undesirable (for example, if at this moment the hoster's restart script is running, an incident with a reboot may occur during fsck).
b) reboot.
If the failure was mild (i.e. just a stop in service), then a light fsck and all is well. If not, depending on circumstances. You should not agree with everything with fsck (without a preliminary backup).
The advice above about smartmon is not relevant - virtual disks do not have smart.
In the future, the panic option in the mount option might help. In this case, in case of disk errors, the system will go into a panic (i.e., into a reboot).

A
Alexey Sundukov, 2012-05-06
@alekciy

Nothing inside the VPS. Yes, and what's the point? All the same, poros can only be solved from the host machine.

P
Pilat, 2012-05-06
@Pilat

Readonly could be for reasons caused by your system, and in case of problems with the storage system at the host. You should have described this situation to the support service, and did not engage in amateur activities if you cannot distinguish between these two options. Moreover, you should not give stupid, in general, advice to check the disks.
To check the disk status, there are smartmontools, smart-notifier. But in the case of VPS, you still need to understand whether this makes sense. Most likely not, it will only create an extra load.

M
mihavxc, 2012-05-06
@mihavxc

I also have a VPS there, how did you scan the disk for errors?

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question