Overview

Right now, the UI derives time remaining for a running workspace from its ttl. This implies that if a user turns on a workspace that is scheduled to stop 8 hours later, then adjusts their ttl to 4 hours, the UI will present that the workspace will shutdown 4 hours later.

However, the underlying model uses a deadline on the build. The deadline may be stickier than ttl - for example, updating your ttl while your workspace is live may or may not cascade to your build.

Therefore it's best that the FE does the following:

• for a running workspace, compute shutdown time from build deadline
• for a non-running workspace, compute shutdown time from workspace ttl

Why It Matters

We want to accurately present when a workspace will shutdown, and not have a mismatch.