Red Hat Bugzilla – Bug 90064
When ETH0 device is trusted, NFS mounts fail when RHL 7.3 clients try to mount them from /etc/fstab while booting.
Last modified: 2007-04-18 12:53:25 EDT
From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20021003 Description of problem: When booting up RHL 7.3 machines which mount NFS shares exported from a RHL 9 machine. The mount fails with "mount: RPC: Port mapper failure - RPC: Unable to receive". After the machines have booted, mount -a will succeed most of the time. When it doesn't succeed, restarting portmapper, or even pinging the NFS server and retrying will usually succeed. When it fails on subsequent mount -a requests it fails with the same RPC port mapper error. Running service iptables stop on the Server fixes the problem: 7.3 clients can then mount the NFS mountpoints during bootup. The port mapper request which goes to the eth0 device, which is set up to be trusted, is being blocked. No lines appear in /var/log/messages on the NFS server machine which they normally do if it receives a request. Version-Release number of selected component (if applicable): iptables-1.2.7a-2 How reproducible: Always Steps to Reproduce: 1. Set up RHL 9 with a valid export on a machine which has a trusted ethernet interface. 2. Add a line in /etc/fstab on a 7.3 client (haven't tried this on another 9 system) to mount the export set up on 1 over the trusted interface. 3. Mount the exported filesystem to ensure that it works. 4. Reboot the 7.3 machine. Actual Results: Mounting fails with the error: mount: RPC: Port mapper failure - RPC: Unable to receive Dropping ipchains on the client machines had no effect on the boot up mount request. Expected Results: 7.3 machine mounts the filesystem successfully. This is the result if ipchains is stopped on the NFS server. Additional info: HW: Dell PowerEdge 2650, Dual Xeon 2.0 GHz, 512 MB RAM, 5 73 GB SCSI drives on PERC 3 RAID 5 array. SW: RHL 9 with all updates applied. /etc/sysconfig/iptables will be attached. I can also upload /etc/fstab (from the clients) and /etc/exports (from the server) if desired. ETH0: BOOTPROTO=STATIC IPADDR=10.2.2.2 NETMASK=255.255.255.240 ETH1: BOOTPROTO=dhcp I have tried a variety of things to get this to work. Please refer to this post on the Shrike mailing list for more information: https://listman.redhat.com/pipermail/shrike-list/2003-April/003320.html
Created attachment 91459 [details] /etc/sysconfig/iptables This is the promised /etc/sysconfig/iptables file.
Please verify this with a newer version of Red Hat Enterprise Linux or Fedora Core and reopen it against the new version if it still occurs. Closing as "not a bug" for now.