?
Solved

Stale NFS between LINUX and Solaris x86

Posted on 2011-03-08
7
Medium Priority
?
841 Views
Last Modified: 2013-12-21
I have a serious issue with NFS mounts going stale between some RHEL servers and a Solaris 10 x86 server.

I am presenting a number of NFS shares from a Solaris 10 x86 server and mounting them on a series of RHEL servers.  Occasionally I need to take the Solaris server offline for software updates, this causes a large number of the LINUX servers to report that the NFS mounts have gone stale.  There are around 400 RHEL servers in total and manually refreshing the mount points is a seriously time consuming task.

Does anyone have any insight as to why the mounts go stale and what, if any, automation could be put in place to re-connect the shares to save my UNIX managers from manually re-connecting each of the mounts.
0
Comment
Question by:phil_howell
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
7 Comments
 
LVL 12

Expert Comment

by:upanwar
ID: 35068101
SInce you have your NFS server on Solaris box so when it goes down then your mounts go stale. The good way for this is use autofs in Linux to mount NFS share.
0
 
LVL 22

Expert Comment

by:blu
ID: 35069257
Of course, while the server is down it is inaccessible. When it comes back up, the NFS handles the clients have should work again, unless the configuration of the server has changed such that the exported file systems have changed device numbers. That shouldn't usually happen unless you are mounting and
unmounting local file systems on the server.  Are you adding new storage?
0
 

Author Comment

by:phil_howell
ID: 35069649
Thanks guys - my hands on team are just checking whether we're using 'autofs' at the moment and looking at the impact of enabling might be if we don't have it.

Blu - we did have issues with the device numbers changing in the past but my storage and UNIX team tell me that they've eliminated that issue now.
0
Free Backup Tool for VMware and Hyper-V

Restore full virtual machine or individual guest files from 19 common file systems directly from the backup file. Schedule VM backups with PowerShell scripts. Set desired time, lean back and let the script to notify you via email upon completion.  

 
LVL 38

Expert Comment

by:yuzh
ID: 35076271
use automount for NFS clients, also check the /etc/default/nfs in your Solsris box, you might need to modify the value for NFSD_SERVERS (Max number of concurrent NFS request), the default value is 16.

you can increase to 32, 64 or 128,  if you have more than one CPU you can bump that up to 256, but try 32 first.

also check you Linux NFS client version, the Solaris 10 x 86 is running is running on v4 by default.

If you want the NFS server run on v3, you need to set
NFS_SERVER_VERSMAX=3

If you modify any value in /etc/default/nfs , you need to restart nfsd (the NFS server daemon).

also you need you make sure you have proper configure the NAME service to make sure the NFS server and client can know each other, (eg use, DNS, /etc/hosts, NIS, LDAP etc)
0
 
LVL 6

Accepted Solution

by:
JRoyse earned 1500 total points
ID: 35082888
Automount is a good idea.

If this is a critical filesystem you could try mounting the clients with a "hard" mount.  You probably are mouting the clients with a "soft" mount.  Sometimes this hangs the client for a bit while the server gets restarted.

You could also build a script to run from cron on the clients to check the mount status and remount the clients.

You could also think about using a NAS/SAN for the mountpoint if this server gets rebooted a lot.
0
 
LVL 38

Expert Comment

by:yuzh
ID: 35089598
You also need to make sure that the NFS server is not overloaded.
0
 

Author Closing Comment

by:phil_howell
ID: 35231327
Thanks all
0

Featured Post

Why Off-Site Backups Are The Only Way To Go

You are probably backing up your data—but how and where? Ransomware is on the rise and there are variants that specifically target backups. Read on to discover why off-site is the way to go.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Determining the an SCCM package name from the Package ID
In this blog, we’ll look at how improvements to Percona XtraDB Cluster improved IST performance.
This video shows how to set up a shell script to accept a positional parameter when called, pass that to a SQL script, accept the output from the statement back and then manipulate it in the Shell.
In this video, Percona Director of Solution Engineering Jon Tobin discusses the function and features of Percona Server for MongoDB. How Percona can help Percona can help you determine if Percona Server for MongoDB is the right solution for …
Suggested Courses
Course of the Month7 days, 19 hours left to enroll

765 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question