Researvation approved, but nodes unavialable, again

24 views
Skip to first unread message

Poan Pan

unread,
May 6, 2021, 8:55:12 PM5/6/21
to cloudlab-users
I reserved 2 r7525 in Clemson, but failed to start the experiment with 2 r7525: https://www.cloudlab.us/status.php?uuid=62e4ad3c-aecc-11eb-b1eb-e4434b2381fc.

From the reservation page, it seems there is 1 r7525 occupied by my project Cascade, but as far as I know there are no other experiments using r7525.

iShot2021-05-07 08.52.53.pngiShot2021-05-07 08.52.36.pngiShot2021-05-07 08.52.23.png

Poan Pan

unread,
May 6, 2021, 10:32:26 PM5/6/21
to cloudlab-users
content in https://www.cloudlab.us/status.php?uuid=62e4ad3c-aecc-11eb-b1eb-e4434b2381fc,I paste the screenshot in case this failed experiment will be garbage collected later.

iShot2021-05-07 10.27.57.png

Mike Hibler

unread,
May 6, 2021, 10:46:24 PM5/6/21
to cloudla...@googlegroups.com
Reservations are per-project, so if there is already one in use by your
project, you can only allocate one more.

If there were other free (unreserved) r7525 nodes, you might be able to
start your experiment, but there are none.

On Thu, May 06, 2021 at 05:55:12PM -0700, Poan Pan wrote:
> I reserved 2 r7525 in Clemson, but failed to start the experiment with 2
> r7525: https://www.cloudlab.us/status.php?uuid=
> 62e4ad3c-aecc-11eb-b1eb-e4434b2381fc.
>
> From the reservation page, it seems there is 1 r7525 occupied by my project
> Cascade, but as far as I know there are no other experiments using r7525.
>
> iShot2021-05-07 08.52.53.pngiShot2021-05-07 08.52.36.pngiShot2021-05-07
> 08.52.23.png
>
> --
> 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 on the web visit https://groups.google.com/d/msgid/
> cloudlab-users/d56e8127-a3e0-4028-bc86-a42a7366a27fn%40googlegroups.com.




Poan Pan

unread,
May 6, 2021, 11:31:22 PM5/6/21
to cloudlab-users
It's clear now.
It turns out that a member in my project starts experiments using small-lan profile without parameterize it, so the experiment in Clemson just occupies one r7525 kind of silently.

Thanks for your response!

Reply all
Reply to author
Forward
0 new messages