Review:
In the last article “NFS problem diagnosis” and you share some common NFS error information, and a similar error after how to diagnose, for this part of the content I particularly emphasize that there is time interested friends must see, or very useful, after all, there is a good saying, the knife is not mistakenly cut wood work. Article link: NFS problem diagnosis
To know the new:
Today we are going to talk about NFS file system links. Why do we want to talk about this? Because some time ago in my operation and maintenance work just out of the NFS file system link related issues. I’m not going to go into the details, but I’m going to give you the final rule of thumb: when mounting NFS, try not to mount it to the root directory. In order to solve this big problem, we did a link to the directory where NFS was mounted, which solved the problem. Here is an example of how to link to NFS.
Experimental instructions: the original mounted in the client /etc/taiji directory NFS file system can be unmounted, and then for /etc/taji do a link to /etc/gufen, and the NFS remounted to the link directory.
1. Export directory /jisuanji on the NFS server
2. Mount the vm to the client
-o bg,intr, and hard are included here because I did not register the NFS filesystem in /etc/filesystems.
3, do links in front of the source directory is the link behind
Do not create gufen in /etc in advance, this command will automatically create the last layer of directory
4. Uninstall /etc/taiji
5, mount to the new directory /etc/gufen
6, validation,
Test /etc/gufen wenjian1 and /etc/taiji wenjian1 and server end of jisuanji wenjian1 is all synchronized
That is, when I operate in the linked directory, the changes are synchronized to the other directories.
This method of linking can be used not only for file system links, but also for common directory links. In fact, the above mentioned NFS file system link is not to do directory links
That’s all we have to say about NFS, and we won’t talk about anything new in the near future