Node boots with wrong image

[john.ou...@gmail.com]

I recently started a large experiment for Homa testing:

https://www.cloudlab.us/status.php?uuid=7e32dff9-3177-4560-8182-36971d49e4e1#

For some reason, node7 of this experiment (er020) was loaded with the wrong image. This is evident by looking in the /boot directory, which contains kernel version 5.15.0; all of the other nodes have the correct kernel version, which is 6.17.8.

I'm going to reload this node with the correct kernel version, but I'll wait a couple of hours to do that in case you'd like to take a look at it. If you decide *not* to take a look at it, can you let me know and I won't wait to reload it?

-John-

[John Ousterhout]

I just noticed that nodes 9, 34, 36, and 41 also have the same problem.

And, in addition, there is a problem with node2 where I cannot ssh to it from node0. I can ssh to it, either from inside CloudLab or outside, using its "full" address (er...@utah.cloudlab.us). I tried power-cycling the node but that did not help. Can you help fix this?

Thanks.

-John-

--
You received this message because you are subscribed to the Google Groups "cloudlab-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cloudlab-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cloudlab-users/3e9d4082-cd26-4cdc-a1ef-c9ec9629c21an%40googlegroups.com.

[Mike Hibler]

I am looking at this. I know what happened just taking a quick look to see
if I can figure out why.

[Mike Hibler]

Can I go ahead and reload all of your nodes? I need to use an option that
is not available via the web interface. I will look at node2 shortly...

> CAGXJAmzCNeQXmJ3BkUSf1Yr_W71Miv_%3DkD%3DQevaiidgSnjrJXA%40mail.gmail.com.

[John Ousterhout]

Go ahead and reload; let me know when you are done?

-John-

To view this discussion visit https://groups.google.com/d/msgid/cloudlab-users/20260706175632.GF79072%40flux.utah.edu.

[Mike Hibler]

They are all reloading now. Give it another 10 minutes and they should be done.
An old version of our disk loader was being used and was incorrectly
identifying which disk got loaded with your image.

> > CAGXJAmzCNeQXmJ3BkUSf1Yr_W71Miv_%3DkD%3DQevaiidgSnjrJXA%40mail.gmail.com.
>
> --
> You received this message because you are subscribed to the Google Groups
> "cloudlab-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cloudlab-user...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/
> cloudlab-users/20260706175632.GF79072%40flux.utah.edu.
>
> --
> You received this message because you are subscribed to the Google Groups
> "cloudlab-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to cloudlab-user...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/cloudlab-users/

> CAGXJAmxum7vjNYv99cR9ztukKQF6DP3dvjjZNBd-eXCJ-Vf_dQ%40mail.gmail.com.

[John Ousterhout]

The nodes are back and they all look good, except for node2 which is still unreachable.

Thanks!

-John-

To view this discussion visit https://groups.google.com/d/msgid/cloudlab-users/20260706182500.GG79072%40flux.utah.edu.

[Mike Hibler]

Sorry that took so long, but the port is back in operation again.

> CAGXJAmwZZ42imKjr08%3D15J%3D5hpKE45ouqMbbWU2brJmsd5LndQ%40mail.gmail.com.

[John Ousterhout]

No problem; thanks for fixing it.

-John-

To view this discussion visit https://groups.google.com/d/msgid/cloudlab-users/20260706215842.GR39666%40flux.utah.edu.