LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: Load Balancing NFS

To: sguercio@xxxxxxxxxxxxxxx,<lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: Re: Load Balancing NFS
From: Matthew Crocker <matthew@xxxxxxxxxxx>
Date: Wed, 11 Jun 2003 16:26:45 -0400

Are you forwarding packets going to the VIP:Portmap to the LVS IP:portmap? Start up ethereal on the external interface of the LVS director and see what packets come in when the client tries to mount,.

From /etc/services I see

sunrpc          111/tcp         portmapper      # RPC 4.0 portmapper TCP
sunrpc          111/udp         portmapper      # RPC 4.0 portmapper UDP
nfs             2049/tcp        nfsd
nfs             2049/udp        nfsd

When I run rpcinfo against my NFS server (Network Appliance)  I see

[matthew@rmc1 etc]$ /usr/sbin/rpcinfo -p 192.168.12.2
   program vers proto   port
    100011    1   udp    608  rquotad
    100021    4   tcp    607  nlockmgr
    100021    3   tcp    607  nlockmgr
    100021    1   tcp    607  nlockmgr
    100021    4   udp    606  nlockmgr
    100021    3   udp    606  nlockmgr
    100021    1   udp    606  nlockmgr
    100024    1   tcp    605  status
    100024    1   udp    604  status
    100005    3   tcp    603  mountd
    100005    2   tcp    603  mountd
    100005    1   tcp    603  mountd
    100005    3   udp    602  mountd
    100005    2   udp    602  mountd
    100005    1   udp    602  mountd
    100003    3   udp   2049  nfs
    100003    2   udp   2049  nfs
    100000    2   tcp    111  portmapper
    100000    2   udp    111  portmapper

You might need to forward all of these ports through. You will probably need to define a FWMARK rule to bundle them all together and then load balance them with persistence to your real servers.


On Wednesday, June 11, 2003, at 03:37 PM, Salvatore J. Guercio Jr. wrote:

Good afternoon lvs users;

I hope you can help me.

I currently have a LVS setup that consists of 2 directors and 2 realservers. I have the LVS setup to serve http right now and It works great. I used the Ultramonkey software packages for RedHat Linux 7.3. The topology I am using the High Capacity High Availability and Load Balancing topology.

The realservers are each connected to an IBM FastT500 fibre channel SAN and I need to export a filesystem on this SAN to 4 clients. The filesystem on the SAN will be GPFS or some parallel filesystem. Has anyone set up ldirectord to export NFS filesystems? I have setup the ltdirectord.cf file to look like this:

virtual=10.9.1.21:2049
       fallback=127.0.0.1:2049
       real=10.9.1.40:2049 gate
       real=10.9.1.41:2049 gate
       service=none
       scheduler=rr
       persistent=600
       protocol=udp

I get the following from   /sbin/ipvsadm -L -n

IP Virtual Server version 1.0.7 (size=65536)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn UDP 10.9.1.21:2049 rr persistent 600 -> 10.9.1.41:2049 Route 1 0 0 -> 10.9.1.40:2049 Route 1 0 0 -> 127.0.0.1:2049 Local 0 0 0 UDP 10.9.1.40:2049 rr persistent 600
TCP  10.9.1.21:80 rr
-> 10.9.1.41:80 Route 1 0 0 -> 10.9.1.40:80 Route 1 0 0 -> 127.0.0.1:80 Local 0 0 0

I can mount the exported filesystem directly from the client, but when I try to mount it from the VIP I get a "mount: RPC: Program not registered" error message.

Is there something (or a bunch of things) that I am missing? Help would be greatly appreciated!

Thank You,

Sal

--
Salvatore J. Guercio Jr.            http://www.ccr.buffalo.edu
Sr. Programmer Analyst              sguercio@xxxxxxxxxxx
Center for Computational Research
9 Norton Hall
Amherst, NY 14260
(716)645-6500


_______________________________________________
LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to http://www.in-addr.de/mailman/listinfo/lvs-users


<Prev in Thread] Current Thread [Next in Thread>