🚀 go-pugleaf

In article <ignacij.712354787@meishan.animal.uiuc.edu> ignacij@meishan.animal.uiuc.edu (Ignacy Misztal) writes:
>From: ignacij@meishan.animal.uiuc.edu (Ignacy Misztal)
>Subject: Re: NFS deamon trap D in tar
>Date: 28 Jul 92 20:19:47 GMT

>Below is the setup in which tar on the Sun was tring to back up 
>the OS/2 files.

>486/33 with 16 M, OS/2 2.0 GA, IBM TCP/IP 1.2.1
>NFSD was started by starting PORTMAP and NFSD with no parameters

>Sun SS1, SunOS 4.1.1, mount line:
>misz:c:/ /home/ignacij/os2c nfs rw,intr,bg,wsize=4096,rsize=4096 0 0


>tar from PC to Sun was started by:
>tar cvf backup os2c

>After copying about 20 files, NFSD quit, printing trap 000d. There was
>no damage to any window. 

>Being unsuccessful, I tried to use rcopy from the PC. It eventually succeded,
>but its pattern of operation was very erratic: 5-20 seconds of seemingly
>high speed copying, followed by a 5-20 s standstill. During that standstill,
>vt100 screen appeared frozen.

>I needed to back up my files to transfer files from FAT to HPFS. Perhaps I  
>should have used xcopy from the PC? 
>----------------------------------------------------------------------
>Ignacy Misztal                  E-mail: ignacy@uiuc.edu
>University of Illinois          Phone: (217) 244-3164
>1207 W. Gregory Dr.
>Urbana, IL 61801
>----------------------------------------------------------------------

For what its worth, here are my $0.02.

I had TCP 1.2 running on OS/2 1.3.  I exported filesystems that were mounted 
on a Sun Sparc IPC.  The performance was very erratic.  Fast spurt, lon 
wait...

Someone told me to sync the time on the PC to the time on the Sun.  When I 
did this, the performance improved dramtically and there were no more 
spurts.

Their reasoning was stale file  handles.  Experts?  Chime in.

Eric
germann@austin.onu.edu