root file system mounted via NFS - retry?

root file system mounted via NFS - retry?

[Randy Smith]

Hello,

I have a general question regarding having a root file system mounted 
via NFS.  Our system consists of a linux PC that acts as the NFS server 
and two embedded ppc boards running linux that mount their file systems 
via NFS.  The problem is that the embedded boards boot much faster than 
the PC, and when they attempt to mount the root file system, the NFS 
server is not up yet.  The embedded boards hang and it takes quite a 
long time before the watchdog timer reboots them.  The next time round, 
they come up just fine.
What I would like for them to do is keep trying until the NFS server 
appears and then continue to boot.

We are running DENX's ELDK 2.4.25 kernel on the ppc boards and I know 
that for some linux releases, one may add values for timeo and retrans 
on the boot command line in the NFSROOT= command.  Can anyone tell me if 
they are valid for this release of the kernel?  Or if there is another 
way to accomplish this?

Thanks,

-Randy Smith
Software Engineer
ImageMap, Inc.

Re: root file system mounted via NFS - retry?

[Jarno Manninen]

On Wednesday 12 April 2006 15:35, Randy Smith wrote:

Hi,

If your system configuration is static why not just sleep in the bootloader 
for a while? U-boot has sleep command.

- Jarno

> Hello,
>
> I have a general question regarding having a root file system mounted
> via NFS.  Our system consists of a linux PC that acts as the NFS server
> and two embedded ppc boards running linux that mount their file systems
> via NFS.  The problem is that the embedded boards boot much faster than
> the PC, and when they attempt to mount the root file system, the NFS
> server is not up yet.  The embedded boards hang and it takes quite a
> long time before the watchdog timer reboots them.  The next time round,
> they come up just fine.
> What I would like for them to do is keep trying until the NFS server
> appears and then continue to boot.
>
> We are running DENX's ELDK 2.4.25 kernel on the ppc boards and I know
> that for some linux releases, one may add values for timeo and retrans
> on the boot command line in the NFSROOT= command.  Can anyone tell me if
> they are valid for this release of the kernel?  Or if there is another
> way to accomplish this?
>
> Thanks,
>
> -Randy Smith
> Software Engineer
> ImageMap, Inc.
>
> _______________________________________________
> Linuxppc-embedded mailing list
> Linuxppc-embedded@ozlabs.org
> https://ozlabs.org/mailman/listinfo/linuxppc-embedded

Re: root file system mounted via NFS - retry?

[Brent Cook]

On Wednesday 12 April 2006 07:35, Randy Smith wrote:
> Hello,
>
> I have a general question regarding having a root file system mounted
> via NFS.  Our system consists of a linux PC that acts as the NFS server
> and two embedded ppc boards running linux that mount their file systems
> via NFS.  The problem is that the embedded boards boot much faster than
> the PC, and when they attempt to mount the root file system, the NFS
> server is not up yet.  The embedded boards hang and it takes quite a
> long time before the watchdog timer reboots them.  The next time round,
> they come up just fine.
> What I would like for them to do is keep trying until the NFS server
> appears and then continue to boot.
>

Sorry to try the obvious, but could you make the firmware's auto-boot timeout 
longer? I know that u-boot and several others support this.

Also, you could try, instead of mounting root NFS directly from the kernel, an 
initrd or embedding an initramfs that does the mount using a script. They you 
could retry as much as you wanted without having to teach the kernel 
anything.

Or, just hook up some GPIOs from your PC to the reset lines on those PPC's. 
Then reset them once your PC is up.

Or rather than running the PPC kernel from flash, tftp it to RAM from your 
PC - that way your PPCs can't even load the kernel until the PC is up 
(assuming your firmware knows how to do this and handles faults gracefully.)

 - Brent

Re: root file system mounted via NFS - retry?

[Randy Smith]

Brent Cook wrote:
> On Wednesday 12 April 2006 07:35, Randy Smith wrote:
>   
>> Hello,
>>
>> I have a general question regarding having a root file system mounted
>> via NFS.  Our system consists of a linux PC that acts as the NFS server
>> and two embedded ppc boards running linux that mount their file systems
>> via NFS.  The problem is that the embedded boards boot much faster than
>> the PC, and when they attempt to mount the root file system, the NFS
>> server is not up yet.  The embedded boards hang and it takes quite a
>> long time before the watchdog timer reboots them.  The next time round,
>> they come up just fine.
>> What I would like for them to do is keep trying until the NFS server
>> appears and then continue to boot.
>>
>>     
>
> Sorry to try the obvious, but could you make the firmware's auto-boot timeout 
> longer? I know that u-boot and several others support this.
>   
We are using U-boot and yes, I can set the delay.  This does work around 
the problem, but I was hoping for a more 'elegant' solution.
> Also, you could try, instead of mounting root NFS directly from the kernel, an 
> initrd or embedding an initramfs that does the mount using a script. They you 
> could retry as much as you wanted without having to teach the kernel 
> anything.
>   
Another possibility.  Thanks!
> Or, just hook up some GPIOs from your PC to the reset lines on those PPC's. 
> Then reset them once your PC is up.
>   
This is not possible with our current design, but it would solve the 
problem.
> Or rather than running the PPC kernel from flash, tftp it to RAM from your 
> PC - that way your PPCs can't even load the kernel until the PC is up 
> (assuming your firmware knows how to do this and handles faults gracefully.)
>   
The way our system was initially designed, it worked this way.  We are 
modifying the design to remove the dependency on NFS by embedding the 
boot loader, kernel and root file system on the flash.  We aren't quite 
there yet so in the meantime we moved the kernel to the flash and now we 
are seeing this problem.

And in fact, the problem may not be as I described.  I have been running 
some other tests and I am getting indications that the problem may be in 
the hardware that our ppc boards are talking to.  I stripped a ppc board 
down to just the computer part and hooked up a serial line to it.  
Cold-boot and the stripped down ppc reached the mount nfs part and 
waited for a little while, and then continued on, so I am discounting 
the theory that the NFS mount was failing due to timeout.  The other ppc 
board hung with the same behavior, but I cannot put a serial line on it.

The question remains though, does the NFS timeo and retrans parameters 
on the nfsroot kernel command line apply to the 2.4.25 kernel supplied 
by DENX' ELDK?

Thanks,

-Randy
>  - Brent
>
>