# Looker query performance

**URL:** <https://discuss.google.dev/t/looker-query-performance/343672>\
**Category:** Looker\
**Tags:** admin, lookml, schedules-alerts, looks, dashboards\
**Created:** [March 26, 2026, 2:15pm UTC](https://discuss.google.dev/t/looker-query-performance/343672 "2026-03-26T14:15:04Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![ksatish](https://d3byx8b92xw3lq.cloudfront.net/user_avatar/discuss.google.dev/ksatish/32/246618_2.png) [@ksatish](https://discuss.google.dev/u/ksatish)\
**Post date:** [March 26, 2026, 2:15pm UTC](https://discuss.google.dev/t/looker-query-performance/343672/1 "2026-03-26T14:15:04Z")

</div>

Monitor and optimize the looker dashboards and queries to reduce load times. So, I need to create the alerts.

1. Looker query running beyond 5 minutes
2. Looker dashboards with queries running beyond 5 minutes

How can I proceed to achieve this.

---

<div class="post-metadata">

**Author:** ![ruthc](https://d3byx8b92xw3lq.cloudfront.net/user_avatar/discuss.google.dev/ruthc/32/562_2.png) [@ruthc](https://discuss.google.dev/u/ruthc)\
**Post date:** [March 26, 2026, 3:19pm UTC](https://discuss.google.dev/t/looker-query-performance/343672/2 "2026-03-26T15:19:43Z")

</div>

5 minutes seems an insane amount of time to wait for a query. I set my threshold at 10 seconds for monitoring the health of our Looker instance. You can use the ‘History’ System Activity Explore to get data on query runtimes.

- The query below shows how many distinct queries have been executed per dashboard (including LookML dashboards), and buckets query into runtimes tiers, based on 5-second increments.
- Personally, I would create my own buckets using the field history.runtime and bucket into \<5, 5-10, 11-20, 20-30, \>30. And apply some conditional formatting to the results.
- Note that you probably want to filter out the ‘undefined’ tier. This will refer to queries that didn’t generate a runtime, such as cached results.

https://\<  
your\_looker\_instance\_url\>/explore/system\_\_activity/history?fields=history.real\_dash\_id,history.query\_run\_count,history.runtime\_tiers\_5&pivots=history.runtime\_tiers\_5&fill\_fields=history.runtime\_tiers\_5&filter\_expression=%28matches\_filter%28%24%7Bdashboard.id%7D%2C+%60NULL%60%29+AND+matches\_filter%28%24%7Bhistory.real\_dash\_id%7D%2C+%60-NULL%60%29%29+OR+%28matches\_filter%28%24%7Bdashboard.id%7D%2C+%60not+null%60%29+AND+matches\_filter%28%24%7Bdashboard.moved\_to\_trash%7D%2C+%60No%60%29%29&sorts=history.runtime\_tiers\_5,history.real\_dash\_id+0&limit=500&column\_limit=50&vis=%7B%22show\_view\_names%22%3Afalse%2C%22show\_row\_numbers%22%3Afalse%2C%22transpose%22%3Afalse%2C%22truncate\_text%22%3Atrue%2C%22hide\_totals%22%3Afalse%2C%22hide\_row\_totals%22%3Afalse%2C%22size\_to\_fit%22%3Atrue%2C%22table\_theme%22%3A%22white%22%2C%22limit\_displayed\_rows%22%3Afalse%2C%22enable\_conditional\_formatting%22%3Afalse%2C%22header\_text\_alignment%22%3A%22left%22%2C%22header\_font\_size%22%3A%2212%22%2C%22rows\_font\_size%22%3A12%2C%22conditional\_formatting\_include\_totals%22%3Afalse%2C%22conditional\_formatting\_include\_nulls%22%3Afalse%2C%22show\_sql\_query\_menu\_options%22%3Afalse%2C%22show\_totals%22%3Atrue%2C%22show\_row\_totals%22%3Atrue%2C%22truncate\_header%22%3Afalse%2C%22minimum\_column\_width%22%3A75%2C%22series\_cell\_visualizations%22%3A%7B%22history.query\_run\_count%22%3A%7B%22is\_active%22%3Afalse%7D%7D%2C%22table\_enable\_pagination%22%3Afalse%2C%22table\_page\_size\_options%22%3A%2220%2C+50%2C+100%22%2C%22table\_column\_hover\_highlight\_enable%22%3Afalse%2C%22table\_show\_headers%22%3Atrue%2C%22header\_font\_bold%22%3Afalse%2C%22header\_font\_italic%22%3Afalse%2C%22cell\_font\_size%22%3A%2212%22%2C%22cell\_font\_weight%22%3A%22%22%2C%22cell\_font\_style%22%3A%22%22%2C%22cell\_text\_alignment%22%3A%22%22%2C%22table\_custom\_border\_enable%22%3Afalse%2C%22table\_custom\_border\_width%22%3Anull%2C%22table\_custom\_border\_color%22%3A%22%23dde2eb%22%2C%22table\_custom\_border\_style%22%3A%22solid%22%2C%22conditional\_formatting%22%3A%5B%5D%2C%22hidden\_pivots%22%3A%7B%7D%2C%22type%22%3A%22looker\_grid%22%2C%22defaults\_version%22%3A1%7D&filter\_config=%7B%7D&origin=share-expanded

---

<div class="post-metadata">

**Author:** ![ksatish](https://d3byx8b92xw3lq.cloudfront.net/user_avatar/discuss.google.dev/ksatish/32/246618_2.png) [@ksatish](https://discuss.google.dev/u/ksatish)\
**Post date:** [April 1, 2026, 11:50am UTC](https://discuss.google.dev/t/looker-query-performance/343672/3 "2026-04-01T11:50:23Z")

</div>

How can we achieve this in live results

---

<div class="post-metadata">

**Author:** ![widle](https://d3byx8b92xw3lq.cloudfront.net/user_avatar/discuss.google.dev/widle/32/145365_2.png) [@widle](https://discuss.google.dev/u/widle)\
**Post date:** [April 15, 2026, 5:01am UTC](https://discuss.google.dev/t/looker-query-performance/343672/4 "2026-04-15T05:01:41Z")

</div>

Hey @ksatish

Use **System Activity → History Explore in Looker** to create a near real-time alert.  
Filter for `history.runtime > 300` (5 minutes) and `history.created_time` = last 5 minutes.  
Build a Look showing query/dashboard details and exclude cached results (`runtime is not null`).  
Set an **Alert** on this Look with condition: _row count \> 0_.  
Schedule it to run every 5 minutes.  
This gives you practical near real-time detection of slow queries without using APIs.

---

<div class="post-metadata">

**Author:** ![ksatish](https://d3byx8b92xw3lq.cloudfront.net/user_avatar/discuss.google.dev/ksatish/32/246618_2.png) [@ksatish](https://discuss.google.dev/u/ksatish)\
**Post date:** [April 15, 2026, 10:09am UTC](https://discuss.google.dev/t/looker-query-performance/343672/5 "2026-04-15T10:09:03Z")

</div>

Thank you @widle
