Still around. At some point I really need to spend the time to write something worth posting here. Most of my notes end up in a private ticketing system with the goal of cleaning them up and posting publicly, but it’s not working out very well.
One thing I will be working on in the coming year is migrating many older/smaller projects over to GitHub. While SVN is still a very useful tool, GitHub is an amazing platform for collaborative development. I’ve been using it quite a bit in the last few months and I’ve come to really appreciate its usefulness. Where I once saw “forking” a project as harmful to a project, I’ve come to understand that Git doesn’t encourage a different model of thinking or coding, it just opens up some options that were not there before. The concern I had about developers taking the code, going off into a cave and not committing back is probably best handled by policy/development culture than it is by the version control system.
Even when hope is in short supply.
Running into cryptic errors really sucks, and this was certainly no exception. We had a Windows share provided to us that we wanted to mount to an Ubuntu 10.04 LTS server for general storage.
After reading the Ubuntu MountWindowsSharesPermanently wiki page and consulting the smbmount man page I came up with this new (initial)
//192.168.1.5/test-smb1 /mnt/it_hosted cifs credentials=/root/it_hosted_windows_share_mount.conf,iocharset=utf8,sec=ntlm2 0 0
/root/it_hosted_windows_share_mount.conf file contained:
I’ve substituted values of course. One thing to note is that if
example.com were the real domain, I’ve made sure to use
example as the value.
Then came the
mount error (13): Permission denied error. I tried from a Windows box and was able to connect fine, so I knew that the server/share wasn’t the problem.
I then decided to call
mount directly as shown here. After getting that working, I rechecked the syntax used in
/etc/fstab and then finally looked at
What do you know, the old & much despised EOL issue. After changing the DOS EOL to UNIX everything worked. I can’t count the number of times I’ve been bitten by that.
Just in case you also get caught by this, this particular storage array does not have a power button. Instead, once power is cycled (unplug the cords and replug), it will boot up and run internal diagnostics prior to powering back up. It took about 15-20 minutes for it come online again after it was safely shutdown.
If you have other hosts/services that depend on it to be online first, it is probably a good idea to wait for the array to come online first before powering them on, otherwise you’ll have to restart any services that depend on it.
I worked a good bit yesterday and earlier today writing up my experiences working with GNU/Linux Software RAID and I feel I’m starting to get a grasp of how it all goes together.
The last array is rebuilding now with another 9+ hours to go, so tomorrow sometime I hope to finish the first draft of that page and get the server back in operation.