top button
Flag Notify
    Connect to us
      Site Registration

Site Registration

device in use after a reboot in centos

0 votes
337 views

I have a server with an 17tb scsi-storage. In past, the storage has a "jfs"-filesystem. Now i want to create a "ext4"-filesystem. I have update the e2fsprogs from 1.41 to 1.42 (16tb limit >1.41).

Now I have an 17tb-storage as /dev/sda1 with ext4. I can mount this device as /home/ (/etc/fstab /dev/sda1 /home/ ext4 defaults 1 2". Now I start a e2fschk /dev/sda1 (umounted). No error-messages are in
the screen.

If I now reboot the server, the server does not started:

I think thats can be a problem with the e2fsprogs 1.42, now i reinstall the server with the default e2fsprogs 1.41 from CentOS 6.4 and create only a 16tb /dev/sda1 partition with ext4. But if i start a "e2fschk
/dev/sda1" and reboot the server, i have the same message in the boot-screen and the server does not boot.

Why the system thinks that the device still be in use? How can i change this?

posted May 17, 2013 by anonymous

Share this question
Facebook Share Button Twitter Share Button LinkedIn Share Button

2 Answers

0 votes

unbelievable!

after a reboot:
sometimes:
/dev/sda1 1 267350 **********+ ee GPT

sometimes:
/dev/sdc1 1 267350 **********+ ee GPT

The server change the device-name.I'm confused - I will use the UUID in fstab

answer May 17, 2013 by anonymous
---/dev/sdc1

Does dmesg say what /dev/sd[ab] is in this case?
0 votes

I think this is not so uncommon. In my experience this has to do with the initialization of PCIe controller cards. I assume you are using a RAID controller, right? I saw that with Areca RAID-controllers in the
past as well with LSI-controller these days.

I am using UUID for that purpose. Not really convenient, but safe. You probably don't have to change that as soon as it is running, so the strange UUID designators are no big issue.

answer May 17, 2013 by anonymous
Similar Questions
0 votes

My setup is as follows:

Host OS: Debian Wheezy amd64 stock kernel (3.2), virt-manager v0.9.1

Guest VMs: A bunch of Linux based servers - CentOS 6, Debian 6/7, Ubuntu 12.04 (all amd64 with stock kernel). All of them defined using virt-manager GUI interface.

From the virt-manager, I am able to shutdown/reboot the Debian and Ubuntu guest OSs but not the CentOS 6 guests.

For the CentOS 6 guests, I have to resort to "Force off"

I have looked at the syslog and messages log files but do see anything to correlate between host and guest OS.

+2 votes

I am used to traditional update-rc.d et all.

Now I wonder how to add a a script that used to called by init.d (with start/sop ..) to the new "service start xx" regime.

All the tutorials I found talk about how to use update-rc.d..

+2 votes

I'm having an issue getting a C6.6 install to work on a 3 TB dual hard drive system, raid 0. I'm hoping that someone here can help.

So, I install as normal, but then reboot, and it comes to a grub prompt. Going into the system via Linux rescue, I see that most of the files dealing with the kernel haven't been installed.

I asked the maker of the server and he said that they have noticed this happen recently. A solution is to put the kernel files on a thumb drive, and then point the OS to look for them there.

I have yet to try it, but is there a better way to deal with this issue that anyone else has done?

+2 votes

I'm planning to use CentOS 6.x on a handful of LAN servers. So far I've been using Slackware64 14.0 and 14.1 for the job.

I wonder what's the orthodox/recommended way of configuring and iptables firewall with CentOS. I understand there's the system-config-securitylevel-tui NCurses interface which allows defining a basic set of rules. But what about the handful of more advanced rules I have to configure?

Here's an example of an /etc/rc.d/rc.firewall script that I might use with Slackware. It contains mostly basic rules, and a couple of more advanced rules, one to limit SSH access, the other one to redirect HTTP traffic to Squid.

If I want to copy my actual firewall configuration to CentOS, what would be the recommended way? I started from a bare bones minimal CentOS 6.5 installation, so system-config-securitylevel-tui is not even installed.
Is it a good idea to try to configure /etc/sysconfig/iptables by hand? What do you suggest?

+6 votes

Noticing multiple threads named "ImageDecoder" showing up in the output of top (when threads are turned on with "H").

Can someone tell me what it is?

...