cloudlab-users

Hi all, If your experimentation involves RDMA, you will need to make sure that your configuration is

> How large can the repository be? Can its size be around 14-16G? That would be too big. Consider

> Instead of deleting it, we could prefix the name with `deprecated` or `broken` or some other

Hi Admin, We have some reservation pending for approval for a while, it is under project DynamicGPU.

We are still having issues with the 100Gb interfaces on the sm110p nodes. We suspect it might be the

Hi Aleks, Thank you very much for confirming, that is good to know. I will check it again when I

Should be fixed. Thanks for letting us know. On Mon, Dec 15, 2025 at 12:40:22PM -0800, Yuan Liang

same here. cannot get any clemson nodes. but am able to get nodes from wisconsin cluster. On Sunday,

Hi Administrators, Could you help me configure the BIOS on node n0 of experiment https://www.cloudlab

You got lucky because the dataset was still in use by your experiment and thus could not be destroyed

Two xl170 nodes do not appear to be working properly: This first node is hp073: cpupower doesn't

> On Dec 12, 2025, at 2:25 PM, John Ousterhout <john.ou...@gmail.com> wrote: > >

Those nodes are connected to Ethernet-only switches, so Infiniband is not possible as I understand it

Thanks; it seems to be working for me now. -John- On Tue, Dec 9, 2025 at 10:11 AM Leigh Stoller <

FYI, the SSD has been replaced and the node is back in service. On Wed, Dec 03, 2025 at 01:26:43PM -

Hi Leonid, It doesn't really matter, I've scheduled it to go out of service once your

Sure, I will keep the experiments there and hope it would be resolved soon. Thanks for your prompt

Thanks a lot Best On Wednesday, December 3, 2025 at 10:23:23 AM UTC-8 Mike Hibler wrote: I extended

FYI, this problem seems to be repeatable: the last 2 times I have started experiments on the c6620

Thank you! On Tuesday, December 2, 2025 at 10:15:53 AM UTC-5 Mike Hibler wrote: Go ahead and

The permissions on ~dk/.ssh had been set to 777; I set it back to 700, so you should be good to login

Unfortunately not. :-( We automatically reap expired short-term datasets. On Sat, Nov 29, 2025 at 07:

Thanks! On Wed, Nov 26, 2025 at 11:57 Leigh Stoller <lbst...@gmail.com> wrote: Hi. This is

Looks good; thanks. -John- On Tue, Nov 25, 2025 at 12:33 PM Mike Hibler <mi...@flux.utah.edu>

The console proxy was not running for ms0933, but it is now. So, you should be able to login on the

Sorry, meant to do this earlier. Your dataset is valid again. On Thu, Nov 20, 2025 at 06:03:08AM -

Hi Admin, I have one dataset which fails my extention request. But I still need for some further

Unfortunately, there is not a way at the moment to reserve specific nodes. You can allocate specific

I have given your project access. On Tue, Nov 18, 2025 at 08:35:22AM -0800, lam...@gmail.com wrote:

Sorry that took so long. Should be fixed now. Assuming you are talking about the Utah cluster... On

Hi Sumit, Just to update you, I wasn't able to get it back online tonight. It should get some

I figured out how to do the Numa Nodes Per Socket changes myself, so a change isn't necessary

Thank you, Mike. I appreciate the update. I hope the issue can be resolved soon. Best regards, Mirco