Issue Unable to use plesk installer

[jonny_alex]

I wanted to access my webmail and ended up with the error Serverfehler: UID SORT: Internal error occurred. Refer to server log for more information. [2020-08-31 14:03:10] (0.017 + 0.000 + 0.016 secs).

Then after some Research i wanted to reinstall my dovecot installation.

Now if i want to use the plesk installer my server says Opening of the file /root/parallels/pool/PSA_18.0.26_2994/release.inf3.tmp for writing: Input/output error failed.

What happend in my server and how can i fix this?

 

[jonny_alex]

UP??

 

[IgorG]

Opening of the file /root/parallels/pool/PSA_18.0.26_2994/release.inf3.tmp for writing: Input/output error failed.

Make sure that you have enough free disk space on this partition, that you have correct permissions.

 

[jonny_alex]

Thanks for your help.

After trying to use the rapir tool the following message appeared:

Is there any problem i can solve on my own?

 

[mow]

/root/parallels/pool/PSA_18.0.26_2994/release.inf3.tmp

• It seems you are using Plesk Obsidian, not Onyx.
• Did you check disk space, like Igor wrote?
• What OS are you running?

 

[jonny_alex]

Yes there is

54 362.18 MB free

38.71%

and

Plesk Obsidian
Version 18.0.27, zuletzt aktualisiert: 16. Mai 2020 14:56:33

 

[jonny_alex]

/root/parallels/pool/PSA_18.0.26_2994/release.inf3.tmp

This File isn't there the folder is empty

AND: All Mails of my clients are stuck in queue




Using the Console this error appeared:

Checking for installed packages...
Downloading file products.inf3: 100% was finished.
Downloading file pool/PSA_18.0.30_4059/release.inf3: 100% was finished.
Downloading file pool/PSA_18.0.29_3884/release.inf3: 100% was finished.
Downloading file pool/PSA_18.0.28_3654/release.inf3: 100% was finished.
Downloading file pool/PSA_18.0.27_3253/release.inf3: 100% was finished.
Downloading file pool/PSA_18.0.26_2994/release.inf3: 0%

Failed to open the file /root/parallels/pool/PSA_18.0.26_2994/release.inf3.tmp for writing: Input/output error.

This may be caused by a problem with the installation source, your connection or the installer settings.
The installer will quit due to an unrecoverable error.
Would you like to view the current installer settings and how to change them?

 

[mow]

Do you have more than one partition on the system? Is /root on its own partition?
If you run "dmesg", do errors show up at the end?
Input/output error can't be much else than a filesystem being either full or in need of a thorough fscking.

 

[jonny_alex]

The server has only one partition sadly.. I have to try dmesg later..

But first it‘s really urgent to get mails working again... My clients are seriously mad they can‘t answer mails. They arrive but no answer possible

 

[jonny_alex]

I added the screenshots wich show the output of dmesg. Can someone read this and see something noticeable?

 

[jonny_alex]

Any news on it?

 

[mow]

Bad news: Your filesystem is corrupted.

All those EXT4-fs errors on /dev/sda2 aren't good. "Error count since last fsck: 14741" is quite a lot.
And it seems your distro's startup scripts don't run fsck on reboot and just (re)mount the partition.

First have a look at "smartctl -a /dev/sda" to see whether the disk is physically broken or it's "only" filesystem corruption.

If you do not have 0 at the end of any of these lines (you might not have all of them in your output):
5 Reallocated_Sector_Ct 0x0032 100 100 --- Old_age Always - 0
187 Reported_Uncorrect 0x0032 100 100 --- Old_age Always - 0
197 Current_Pending_Sector 0x0012 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0010 100 100 000 Old_age Offline - 0
then your harddisk is broken and needs to be replaced.

Is it a hosted server or do you have physical access to it so you can run a repair system and copy the data with e.g. ddrescue to another disk?
If it is hosted, and smartctl shows the disk is broken, the hoster is responsible for replacing the disk, but you will most certainly get an empty one so you better have a complete and recent backup. It might be too late to create one with plesk now (as it will probably fail on harddisk errors).

 

[jonny_alex]

# smartctl -a /dev/sda -T permissive
smartctl 7.0 2018-12-30 r4883 [x86_64-linux-3.10.0-327.el7.x86_64] (local build)
Copyright (C) 2002-18, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Device Model: harddisk
Serial Number: RHTVHNCFC2B4H2DRNYAX
Firmware Version: F.YBCC70
User Capacity: 107'374'182'400 bytes [107 GB]
Sector Sizes: 512 bytes logical, 4096 bytes physical
Rotation Rate: 7200 rpm
Device is: Not in smartctl database [for details use: -P showall]
ATA Version is: ATA/ATAPI-6, ATA/ATAPI-5 T13/1321D revision 1
SATA Version is: SATA 2.6, 3.0 Gb/s
Local Time is: Thu Sep 17 20:59:39 2020 CEST
SMART support is: Unavailable - device lacks SMART capability.

SMART Disabled. Use option -s with argument 'on' to enable it.
(override with '-T permissive' option)

What does that mean to me what do i have to get ti informations you need?

It's pretty scary since last backups are always done but surely also corrupted.. What's the best way out of this? I can't figure it out since all my clients (10 People) rely on the servers health to communicate to their clients like me aswell..

My support "hosttech" wants SFR. 450/hour for support so i try to recive help from here since i noticed you are well known users.
BIG THANK YOU so far for helping me with your knowledge i appreciate this alot

 

[mow]

You could try if "smartctl -s on /dev/sda" is able to switch on SMART so you can run "smartctl -a /dev/sda" afterwards.

Are backups stored externally?

Your hoster's support should not charge if it's their hardware that failed, but as I wrote they would probably leave you with a blank new disk.

There is still a chance that "only" the filesystem is corrupted and not the disk itself, but it is a risk to run fsck on a remote server. Normally, you'd always copy the whole disk elsewhere with ddrescue, but that's not so easy to do with a remote server without the help of the hoster ...

Steps to do it the hard way:
make sure you have backed up all you can
run the following:
service mysql stop
mount /dev/sda2 -o remount,ro
fsck -p /dev/sda2
if that does not want to run, use
fsck -p -f /dev/sda2
hope the errors are not too horrible
if they are and fsck bails out you can try
fsck -y -f /dev/sda2
but it might ruin the rest of the filesystem, however at this point there's not much else you can do
after the fsck, reboot
shutdown -r now
and hope it comes up again

 

[jonny_alex]

You could try if "smartctl -s on /dev/sda" is able to switch on SMART so you can run "smartctl -a /dev/sda" afterwards.

"smartctl 7.0 2018-12-30 r4883 [x86_64-linux-3.10.0-327.el7.x86_64] (local build)
Copyright (C) 2002-18, Bruce Allen, Christian Franke, www.smartmontools.org

SMART support is: Unavailable - device lacks SMART capability.
=== START OF ENABLE/DISABLE COMMANDS SECTION ===
SMART Enable failed: scsi error badly formed scsi parameters

A mandatory SMART command failed: exiting. "

Are backups stored externally?"

Can't tell i'd have to ask my host

There is still a chance that "only" the filesystem is corrupted and not the disk itself, but it is a risk to run fsck on a remote server. Normally, you'd always copy the whole disk elsewhere with ddrescue, but that's not so easy to do with a remote server without the help of the hoster ...

I'll try and see with my host

 

[jonny_alex]

• service mysql stop

Redirecting to /bin/systemctl stop mysql.service
Failed to stop mysql.service: Unit mysql.service not loaded.

• mount /dev/sda2 -o remount,ro

mount: / is busy

• fsck -p /dev/sda2

fsck von util-linux 2.23.2
/dev/sda2 ist mounted.
e2fsck: cannot continue aborting.

Soo none of this worked sadly... What else would you recommend me?

It's a Vserver just as a detailed info... So do you think it was a user fail (Me) or the host?

 

[mow]

If it is a vserver, this needs to be fixed on the host.