Reduce CPU, memory, and crawl load

Lower local workload without losing the most valuable monitoring.

Last checked 20 August 2026

TechDash renders monitored pages in a local browser, stores compressed HTML, captures screenshots, and runs several lookup queues. A large or frequent portfolio can use noticeable CPU, memory, disk, and network capacity. Reduce the work in stages so the most valuable monitoring remains current.

Lower simultaneous page checks

  1. Open Settings.
  2. Find Monitoring.
  3. Lower Concurrent Crawls by one step.
  4. Let the app run long enough to compare computer load and Queue Status.

Concurrent Crawls controls how many page checks can run at the same time. TechDash calculates a recommended value and a device-specific maximum from CPU count and available RAM. The maximum is capped by CPU cores minus one and RAM. The normal recommendation is two where the hardware supports it, but the interface shows the recommendation for this computer.

Lower concurrency reduces peak browser work. It does not reduce the number of checks due each day, so lowering it too far can increase the waiting queue. Use the lowest value that keeps the computer comfortable without creating a persistent backlog.

Find the high-load websites

Open Queue Status and review Website Pressure. It combines enabled pages, queued pages, slow pages, large stored HTML, and estimated checks per day. Use it to choose where a scope or frequency change will have the most effect.

Also review the separate queue rows. Page concurrency directly affects Page checks, while Site resources, SSL, Domain, DNS, Search Console weekly, Search Console URL inspection, Screenshots, and Icons have their own workers and schedules.

Reduce frequency before removing coverage

The available page intervals are Every 15 minutes, Hourly, Daily, and Weekly. Frequent monitoring multiplies the workload:

  • One page at Every 15 minutes can become about 96 due checks per day.
  • The same page at Hourly is about 24.
  • Daily is about one.
  • Weekly spreads the work across a week.

These are schedule estimates, not guarantees. Sleep, battery state, a locked screen, backlogs, and network availability can delay a run.

Keep short intervals for pages where rapid detection is valuable. Change stable or lower-priority pages to Daily or Weekly. If a whole website needs the same cadence, change its website frequency rather than editing many pages individually.

Reduce monitored scope

Disable or remove low-value pages that duplicate the same template or do not need ongoing evidence. Disabling a page preserves its record while stopping scheduled page checks. Removing it is a deletion action. Pause an entire website when no checks are currently needed and resume it later.

Be selective before removing representative pages. Monitoring fewer pages can reduce the ability to detect template-specific failures, internal-link changes, duplicate patterns, or sitemap membership problems.

Investigate slow or large pages

Website Pressure identifies slow checks and large stored HTML. Large HTML can increase disk use and processing, while scripts, third-party requests, lazy-loaded assets, bot challenges, and slow servers can lengthen browser work.

Confirm that the saved user agent and Basic Authentication settings are correct. A page stuck behind a challenge or login can still consume time without producing useful results. Compare a representative page in a normal browser from the same computer.

Stored raw and rendered HTML is compressed, but compression does not make an unusually large page inexpensive to render. TechDash also saves website screenshots and icons locally.

Understand automatic load reduction

On battery power or while the screen is locked, page-check intervals stretch to about twice the configured interval. This lowers activity without changing your saved frequency. It does not pause all queues and should not replace deliberate tuning for a permanently large portfolio.

After each change, review waiting and active counts plus last-run times. If memory remains high with no active or waiting work, quit TechDash completely, reopen it, and collect the app version, platform, queue state, and affected pages for support.

Related articles