Read waiting jobs, active jobs, and worker counts
Interpret the three core metrics for each scheduler queue.
Last checked 20 August 2026
Every Queue Status lane shows three numbers. Read them together to understand whether work is arriving, running, and draining.
Waiting Jobs
Waiting Jobs is work already admitted to a queue but not yet started. It is not necessarily every task that will eventually become due. Schedulers admit work over time.
For page checks, the scheduler runs a due-work sweep every minute. Automatic admission is capped at the greater of eight waiting jobs or five times the page-check concurrency. With two workers, for example, the automatic waiting cap is 10. Manual Run check now requests bypass that admission cap and can increase waiting work.
Due page checks are ordered with the most overdue work first, then interleaved across websites to provide a fair share. A large site should not consume every newly available slot before another site's overdue page is admitted.
Waiting is green at zero, amber above zero, and red when it reaches the backlog threshold. The backlog threshold is the greater of eight or three times the worker count.
Active Jobs
Active Jobs is work currently being processed by the lane. Active can be lower than workers when there is not enough waiting work, jobs have just completed, or a scheduler is rate-limited.
The number is green when below concurrency. It is amber when all workers are occupied but nothing is waiting. It is red when all workers are occupied and at least one job is waiting.
An active count of zero is healthy when waiting is also zero. It is more concerning when waiting persists. In that case, check whether the app is activated, the website is paused, the computer is offline or behind a captive Wi-Fi sign-in, or a recent error is preventing work from starting.
Workers
Workers is the maximum concurrent jobs allowed for that queue at that moment. It is not a thread, process, CPU-core, or device count.
For Page checks, workers matches Settings > Monitoring > Concurrent Crawls. TechDash constrains that setting to at least one and to a hardware maximum based on CPU cores minus one and installed RAM. The recommended default is two when the hardware allows it.
Other queues have independently configured concurrency and timing. Search Console work, DNS, domain, SSL, site-resource, screenshot, and icon jobs do not inherit the Concurrent Crawls slider. Some deliberately remain low-concurrency or rate-limited to protect external services and local resources.
The Workers number uses the same tone as Waiting Jobs. Red workers therefore does not mean the workers themselves failed. It means waiting work has crossed that queue's backlog threshold.
Recognize common combinations
- Waiting 0, Active 0: the lane is idle.
- Waiting 0, Active below Workers: available work is running without a backlog.
- Waiting 0, Active equals Workers: all capacity is busy, but the queue has caught up.
- Waiting above 0, Active equals Workers: all capacity is occupied and work is queued.
- Waiting rising, Active equals Workers: demand is exceeding current throughput.
- Waiting above 0, Active 0: work is stalled or between starts and needs investigation if it persists.
Do not diagnose a backlog from one snapshot. Queue Status updates every 10 seconds, and short bursts are expected after startup, resume, connectivity recovery, a bulk page add, or manual checks. Watch whether Waiting falls over several refreshes.
If Page checks stays backlogged, reduce low-value page frequency or page count before raising concurrency. Raise Concurrent Crawls only when the computer has spare CPU and memory and the current jobs are not constrained by the network or destination site.