lost connection always
I have two servers, these two servers are constantly disconnected, my other l40s servers are not constantly disconnected in this way, what is the reason for this, these are financial losses for me,...
Sudden 10x Price Increase After Adding Storage
I deployed a VM at $0.01/hr for 200GB storage. After simply adding +100GB storage, the hourly cost unexpectedly jumped to $0.10/hr — a 10x increase. The main issues here: There was no warning or...
Spending rate seems wrong
I'm running a VM which costs 2.19 per hour. 2 hours ago, I had 15 bucks left. Now I have <4. It seems like the billing rate is far higher then shown. Please help me understand the billing rate here...
No instruction given on how to access jupyter in new beta site.
I waisted decent amount of money and time and just exploring how to connect to jupyter lab interface but could not access. Earlier you used to provide clear url to access jupyter lab but now that...
An unexpected error occurred during deployment
when I try creating new GPU usually it is "4090 or 3090", it keeps popping up the same error "An unexpected error occurred during deployment." even though it shows 5 or 6 units available, I tried...
Overcharging , they never replied the email 👎
They charge me as if the server is running 24/7, my balance goes down 15-20 dollars per day even if server is not running. I emailed them and got no replies
Can we please bring back interruptible/spot instances for Tensordock.
Title. Because my H100 used to only cost 1.6 USD per GPU hour (when the standard costed 3.3 usd per GPU hour), but now it's gone up to 2.2 USD per GPU hour standard, and i don't have the option to...
Usage Email Alert
Hi there, It would be very helpful if the system allowed users to set a usage threshold and receive an email alert when that limit is reached. Thank you!
The gpu selection is a barren desert and the lack of public ip servers is distressing
The lack of persistent storage that can be attached to different gpus due to availability, coupled with the overall lack of gpus, as well as the lack of servers with public is as an option turned my...
Metadata Endpoint
Need a way for the instance to find its own tensordock uuid and other metadata