What can I say other than I’m a bit of a Pill*ck! I’m a firm advocate of backing up your data and creating disk image backups, with Windows and Macrium reflect its easy as pie. Windows 10 will backup up your stuff anyway, so there’s no excuse. Unfortunately it doesn’t seem to be as easy to create disk images with Linux (Centos 7) so I haven’t got around to doing a disk image of the “System disk” but do backup the data drive daily. My knowledge of Linux generally is lacking, compared to “Windows” but I tend to use things like Webmin, Putty and WinSCP to help me administer Linux boxes and generally I get by using these tools. In the past, I have administered SCO Unix, Solaris, SUSE, Ubuntu and other Linux Distros but have always struggled with the lack of a decent GUI. Maybe I’m command line phobic 🙂
I’ve found that CentOS 7 is normally rock solid and the only time I’ve had to reboot was on a couple of powercuts where my UPS has let me down, so I haven’t really got around to finding out how to do a disk image, OK that’s enough excuses. Last week, Webmin advised there were about 79 updates so I left it running and went to bed. The updates failed and I was unable to access the CentOS 7 server remotely, mmmm that’s not good! Sitting at the box (in a cold garage) and working interactively, I find that CentOS was in “Emergency Mode”. Some 10 hours of trying everything I know and trying stuff found on the Web has resulted in “mmmm, thats not good” 🙁 Had it been a Windows box, I reckon I would have been up and runing in about an hour!
It seems that my only option now is to do a complete re-install of CentOS 7 and have to set up the SAMBA shares and sort out the DNS server bit, recreate the back ups for the Data and all because I don’t have a disk image. I could blame Webmin or the people that have issued the updates but although the updates should have completed without any issues, the lack of a disk image backup is down to me!
I’ve now found a “Live Disk Image creator”, Mondorescue, and now I’ve got it working by amending the size of the RAM drive, it’s wonderful! It creates ISO files so now I can create disk images of the “system disk” with ease so there’s no excuse for not backing up the system disk.
It’s very true that the backup you regret most is the one you didn’t do!
For the record, I don’t yet know what caused the failure. I tried a clean install on a different HDD but after much swearing and cursing; I found the spare HDD had bad sectors, grrr, spit ,grrrr. Different HDD and all went smooth as a babies butt, except for getting Webmin working and the aforementioned Mondorescue. I’ve done all the upadates, a few at a time and all working nicely. Setting up the Cron Jobs to get rsync to backup the internal data HDD to the ext HDD had a few challanges but got those sorted so I’m pretty much back to where I was a week ago; when everything was running fine.
Suppose I’d better check to see if those ISOs (created by Mondo) do actually work.
All this crap I’ve dealt with the past week would have been considerably reduced if I’d done a disk image backup. It’s a case of “physician heal thy selve” and pill*ck is being being kind 🙂
In my case, the Centos server is purely there to provide network shares. Had it been an essential bit of a business, a P45 would be expected 🙁
Knowledge is power 🙂