Parallel run limits per organization, suite and manual execution

Problem
Teams with many parallel cloud runs often use all of them for automated suites (scheduled runs, CI/CD, API triggers). This causes two issues:

• When someone needs to run a test manually, for example to verify a hotfix or debug a failure, they have to wait until an automated run finishes, because every slot is already taken.

• One large suite can take all available runs and block other suites, so a quick smoke test waits behind a long regression run.

What we’re planning
Concurrency limits on two levels:

1. Organization level: decide how many of your parallel runs automated executions may use at the same time, and keep the rest free for tests started manually from the BugBug app.

2. Suite level: decide how many parallel runs each test suite may use, so you can split your capacity between suites.

Example: an organization with 32 parallel runs sets the automation limit to 28, so 4 runs are always free for manual testing. Within those 28, it sets:

• Full regression suite: max 16 parallel runs

• Smoke tests: max 8 parallel runs

• Nightly checks: max 4 parallel runs

The regression suite can never take the slots the smoke tests need, and manual testing is never blocked.

How it will work

• A new setting in organization settings: “Max parallel runs for automated executions”

• A new setting in each suite’s settings: “Max parallel runs for this suite” (by default no limit, so the suite uses whatever is available)

• Suite limits work inside the organization limit, which always applies

• If the suite limits add up to no more than the organization limit, each suite effectively has its own guaranteed capacity

• Runs above any limit are queued, not rejected

• Manual runs can use any free slot, including the reserved ones

• Works within your current plan, with no change to your subscription or number of parallel runs