OK – I don’t work for Heart so this isn’t an excuse but when it comes to VPSs there is a reason they are being so slow that I am aware of. Applications like WordPress, X-Cart and many others aren’t designed to behave well in a crash situation and strange things can happen.
As far as I can see the database servers went down at Heart before the main servers meaning many applications behaved in a very wierd way and it takes ages to find out what. In one example of mine an error log file ballooned to 40GB making it impossible to restart the server.
I’ve recovered all my clients VPSs now and this is the step by step how to although you may have to mix and match a bit. Some of it may sound technical but anyone can do it.
Disclaimer: I take no responsibility for any loss caused by following this guide, you follow it at your own risk (have to put this in for legal reasons!)
Note – the following may get your website working only for it to fall over again or your site may not work but Plesk will. Please see the last part of this post about repairing your MySQL tables to avoid this.
First open the Heart recovery guide in a new window.
Second you will need the following two programmes:
- puTTy – download here (from the first link on the page)
- WinSCP – download here for this sort of FTP service that allows root access
Restarting in recovery mode
Login to your Heart panel, go to Manage VPS and Hybrid Servers, click on the server you want to work on and look under the ‘Connectivity’ heading for the login data you will need.
Open puTTy, put the IP address in the box marked Host Name (or IP address) and click Open. A black screen will appear with ‘login as:’. Type root and hit ENTER. The system will ask for your password. Type it in (copy and paste doesn’t always work) but be aware that nothing happens as you type – don’t worry, it is going in. Hit ENTER.
Now you can run the fsck – /dev/sda3 command you will have seen in the Heart recovery guide
Still not working? The next step is to see if it is down to MySQL corruption
Step 1-7 in the Heart guide
Open the /etc/my.cnf file (MySQL’s config file) … but how?
Launch WinSCP, enter the IP address as the Host name and click Login. Enter the password from the step above. Now you have root access to the server file structure (right hand panel). You may get logged in anywhere so, in the right hand panel, keep going up a folder (the little yellow folder icon top right with a L shaped arrow pointing up) until you can’t go any further.
Now you can navigate to the etc folder and download the my.cnf file as per Step 1 of the Heart guide by dragging it from the right panel (the server) to the left panel (your computer). Open using notepad or another text editor and edit as per Step 2. Save it and then in WinSCP upload to your server by dragging from the left panel to the right.
Try starting MySQL again – but how?
Back to puTTy enter service mysqld start. You may also want to see this page for a few alternatives to try such as stopping and starting.
Remember to repeat this as per the guide up to step 7.
If this works out remember to delete the changes to the my.cnf file and upload it to your server.
Still not working?
Check for a full disk overload
In WinSCP navigate to var/www/vhosts/ and then the domain name. Then enter the folder httpdocs (double click). Right hand click on each folder and chose ‘Properties’. In the window that opens click ‘Calculate’ to the right hand side of Size.
If you find a directory that seems to be crazy big double click on it and root out the guilty file which is probably an error log. You’re looking for a file which would have filled up all available space on your server – 10 GB+.
You are going to need to delete this so make sure you are confident that it looks like a log file (it should be called something like ‘error_log.txt’ or ‘activity_log.txt’. If you are not 100% sure you can download it to your computer (click and drag to the left panel) just in case as long as you have a very fast internet connection.
Once deleted go back up to the top of this post and repeat. Personally I skipped trying to start the server in recovery and just went straight to restarting the MySQL part in most cases.
Started working but fell over again?
Chances are your MySQL tables are all over the place. If you can now login to Plesk go there and to MySQL admin. Scroll to the bottom of the list of tables, click ‘Check All’ and from the select box ‘Repair table’. Repeat this for each database