Date: Tue, 23 Nov 1999 12:26:07 -0700 From: "Juan Lorenzana" <lorenzaj@agcs.com> To: freebsd-questions@freebsd.org, hackers@freebsd.org Subject: NFS hangs/locks CPU Message-ID: <383AEA4F.ABE8D3DD@agcs.com>
next in thread | raw e-mail | index | archive | help
Hi, I'm running FreeBSD 2.2.8. We have used 2.2.8 to build a telecom application and have had great success deploying other telecom applications on 2.2.8 until now. We are using NFS to distribute binaries and config files to others machines so that we can achieve availability. Here's our problem: I have two machines one acting as a server and the other as a client. The processes are started as follows NFS Server (king-n1) mountd nfsd -u -t 10 rpc.statd NFS Client (king-c1) mount_nfs -T -w 1024 king-n1:/u /u Both machines are Pentium II 450 mhz with 512 megs of ram and 1024 megs of swap. We use an Intel Ethernet Express Pro 100 cards with the fxp0 driver. I turned on debug for the fxp0 driver and see nothing out of the ordinary. Please note that king-n1 is a dual homed. That is why we use the -T option. Otherwise, we can not use NFS to the virtual ip address without installing the NFS_Security_Cookie workaround. Since the nfs_security_cookie workaround has not been tested on 2.2.8, we chose to use the -T option (odd but it allows us to mount a directory from a virtual ip address instead of the real ip address). It is interesting to note that this hang still occurs whether or not we use the real ip address or a virtual ip address. The problem is that under heavy load, we see that the client does a lot of Getattr and Access (nfsstat -w 1). We were able to compile the DDB option into the kernel and when the machine hung,we entered into the debugger and saw that we were stuck in a _write call. The following is what we saw when we did a trace. _boidone(f418d200,f418d200,f22bf000,f22dc600,3) at _boidone 0x2e6 _nfs_doio(f418d200,f22df000,f22dc600,1,f418d200) at _nfs_doio 0x4c5 _nfs_strategy(efbffddc) at _nfs_strategy 0x68 _nfs_writebp(f418d200,1,efbffec4,f0163000,efbffe50) at _nfs_writebp 0x125 _nfs_bwrite(efbff50) at _nfs_bwrite 0x10 _nfs_write(efbffee08,f02564e0,1,efbff94,2) at _nfs_write 0x648 _vn_write(f25d6f80,efbfff34,f22bf000,f02564e0,f22c600) at _vn_write 0x93 _write(f220c600,efbff94,efbff84,0,5167) at _write 0x97 _syscall I am not sure what to do from here. We can only run traffic for about 4 hours before our machine hangs. I suspect that this might be related to a GETATTR or ACCESS nfs problem or a possible tcp/ip bug under heavy load. Can anyone help or point me in the right direction. Moving to 3.x is not possible because we are using a product that we purchased from Sweden. The Swedish product is the base of all our telecom products here at AGCS and it only runs on 2.2.8. Any help is appreciated. Thanks. Regards, -- Juan Lorenzana AG Communication Systems Phoenix, AZ 602-582-7442 lorenzaj@agcs.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?383AEA4F.ABE8D3DD>