TensorDock users flag failed VM restarts

This reads as an operational trust problem, not a one-off bug report: the complaint is about recovery, support latency, and whether paid storage actually protects valuable work.

• –The core issue is VM availability and restoreability, not raw GPU performance.
• –The user claims they kept paying for storage specifically to avoid data loss, which makes the failed startup more damaging.
• –The lack of support response is part of the complaint and amplifies the reliability concern.
• –The post also suggests TensorDock may not offer a simple path to remount an existing disk image onto a new VM.
• –The “40x credits” mention sounds like compensation policy language that is not clearly understood by the user.