CentOS - Sep 2014 - CentOS on VPS File System in read-only mode

Hello All,

First of all I want say thanks for all the info I see every time in this
list.
Second, sorry for my basic english.

I writting this because I'm experiencing an issue with a Centos over a VPS.
Every saturday (I have installed this vps since 3 weeks ago) I check the
status of the server I found that the file system is in read-only mode

#touch test
touch: cannot touch `test': Read-only file system

# mount -o remount /
mount: block device /dev/sda1 is write-protected, mounting read-only

Looking for info in the logs, I don't know how to get the root cause of the
issue. But I can see that the last writing process is marked about the 1:40
hr

#ls -ltr - /var/log
-rw------- 1 root     root     479825 Sep 13 01:20 maillog
-rw------- 1 root     root     229268 Sep 13 01:30 cron
-rw------- 1 root     root     106941 Sep 13 01:34 messages

(now is Sep 13 18hr)

The centos is mounted on a VPS. The VPS server is a Proxmox. The sysadmin
of the VPS told me that there is a backup process running on friday at
midnight.

I suppose that the issue is a very high I/O request on the HD during the
backup process.
( example in vmware
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=51306
)

# uname -a
Linux 2.6.18-371.12.1.el5 #1 SMP Wed Sep 3 16:22:34 EDT 2014 x86_64 x86_64
x86_64 GNU/Linux

There is any kind of workarround for this (make the centos more flexible
about the I/O timeouts or fails)
Or maybe there is something to recommend to the VPS sysadmin to solve this
issue?

more info:
# more /etc/fstab
LABEL=/                 /                       ext3    defaults        1 1
tmpfs                   /dev/shm                tmpfs   defaults        0 0
devpts                  /dev/pts                devpts  gid=5,mode=620  0 0
sysfs                   /sys                    sysfs   defaults        0 0
proc                    /proc                   proc    defaults        0 0
LABEL=SWAP-sda2         swap                    swap    defaults        0 0

#df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda1              19G  3.4G   15G  19% /
tmpfs                 502M     0  502M   0% /dev/shm

The previous times it happened the fsck after a reboot fixed the issue.

Thanks in advance.
PP

On Sat, Sep 13, 2014 at 06:46:40PM -0300, Pablo Parodi wrote:
> Every saturday (I have installed this vps since 3 weeks ago) I check the
> status of the server I found that the file system is in read-only mode
The most common reason for this is a disk that is throwing errors. When this is
the system disk, it inhibits logging the error in /var/log/messages. However,
the error will be logged in other places.

3 places to check:

1) Run the 'dmesg' command, look for errors
1a) (If you had a remote syslog server configured, you won't miss errors
in /var/log/messages because the system disk is read-only)

2) 'cat /dev/vcs' to see if there is anything on the console. (A server
will
ideally have a console log configured to not miss anything.)

3) 'smartctl -a /dev/sda' will show you if the disk has failed its
self-test.
(Ideally you have it configured to email you in this case.)
> I suppose that the issue is a very high I/O request on the HD during the
> backup process.
I have never seen high I/O rates cause a read-only filesystem, except
in the case where a disk error was visible. Linux has no problems with
extremely busy disks; I have clusters with 100s of servers that run their
disks with high I/O requests 24 hours a day for years, with no problems that
were not visible in the logs 1-3.

-- greg

p.s. your English is pretty good!