https://nf-co.re logo
Join Slack
Powered by
Hi, did something change in the GPU compute enviro...
# aws-megatests
j
Hi, did something change in the GPU compute environment? I was able to run the proteinfold modes last week, but now some of them failed when pulling the images with a 
no space left on device
error
r
I haven't touched it
j
I had this error before, but fix it adding 
docker.pullStrategy = 'lazy'
, but now they don’t work even with this flag
r
maybe compare what machine types you submitted to in both cases?
j
It seems that the compute environment of the successful runs has been deleted
🥲 1
Is there any workaround or any possibility of setting again the previous environment? cat eyes Otherwise, I am afraid that #C0351A8S487 test for the release can’t be run sadpanda
r
the environemtns were created the same way. Can you please link the run that worked too?
the only difference I see is the boot size disk. we can try to bump that.
give me a minute
👍🏻 1
j
Thanks a lot!
r
done
j
OK, will try right now
shared a GIF
Everything worked now but a couple of runs 🎉
After checking the two tests that are failing are still having a 
no space left on device
so I have a question did the environment where you increased the boot size not match exactly the previous working environment? And would it be possible to increase the size at least until the proteinfold tests run? please
r
You can compare the two CEs by clicking the "View" button for each. and check the exact configuraiton it was stood up with. I thought only the boot size was different, but maybe you can double check all the fields again
👍🏻 1
j
The successful envirnoment has nothing in the 
Boot disk size
while the new has 50 Gb, if seqera ai is correct, when not provided and fusion 2 enable, the size of the boot disk is set to 100 GB by default, would it be possible to update the environment?
r
can you link the successful environemnt?
j
2 Views
Open in Slack
PreviousNext
Powered by