How to build a PromQL (Prometheus Query Language)

This episode will explain how to build a Promql.
What you are going to learn out of this episode.
- Introduction on the various data types of Prometheus
- The various format of data types to store metrics in Prometheus
- The various way to filter data
- The various operators
- A small tutorial on various queries
Links:
- Tutorial: github.com/isItObservable/Epi...
- Blogpost : isitobservable.io/observabili...
Timestamps
00:19 - Introduction
02:39 - Type of Prometheus metric
09:04 - Prometheus data type
23:36 - Operators
25:53 - Tutorial

[@cramhead]

Your video was way more approachable than the PromQL docs. Thanks so much for sharing your knowledge and insights.

[@isitobservable]

Thanks for you feedback

[@guilhermehubner1991]

Really nice video!! I am finally feeling more confident about promQL! Keep the nice job!

[@isitobservable]

Glad I could help!

[@rshahdelhi]

Very nice tutorial

[@vynguyenhoang5604]

Thank you for a video with a very detail explanation!

[@GustavoRodrigues-le3zw]

Incredible content quality! You will grow a lot! Thanks for the content!!

[@isitobservable]

Thanks for your feedback. Let me know if there are any topics that i should cover.

[@alanmangroo3656]

Great video, very well explained.

[@sunnynew100]

Very nice video. Looking forward to more such content

[@isitobservable]

Thank you, I will produce more content. let me know if there are topics that could help ;-)

[@riskofrains]

Thanks for this lesson!

[@debasisnath9860]

Great! Perfect description of concepts ✨✨✨✨✨✨

[@isitobservable]

Thank you! 😃

[@rajeevkumar-pc4ne]

it is ice breaker video. thanka lot sir..

[@Trojaner]

Thank you, helped me a lot learning Prometheus

[@isitobservable]

Glad I could help!

[@um4r_arf]

thanks after many vids i found a useful one

[@annapureddyravi4054]

Excellent demo..thank you so much

[@isitobservable]

Glad you liked it!

[@stephaniesilva1939]

Thank you so much! I love your contents!

[@isitobservable]

Glad you like them!

[@mauridocarmo7167]

Awesome video. I've learned a lot.

[@isitobservable]

Glad you enjoyed it!

[@vg3348]

thank you for the video. very informative.

[@isitobservable]

Glad it was helpful!

[@egidfertal8240]

thank you very much :) this video was were usefull for me :)

[@isitobservable]

Glad it helped!

[@sumanreddyk7284]

Excellent Work! I was searching for this level of PromQL information on KZbin. When you get a chance, could you share your knowledge on grafana dashboard creation using json.

[@isitobservable]

Thanks for you feedback.I'm adding this topic in my backlog.

[@lunyi2004]

thanks, your video help me a lot.

[@isitobservable]

Glad to hear that!

[@eyuandedenchennel7681]

Awesome vedio thanks a lot!!

[@isitobservable]

Glad you liked it!

[@sunhsiang6644]

Thanks

[@jacoduplessis8167]

Great video thanks

[@isitobservable]

Glad you enjoyed it

[@vissu403]

Really Nice video

[@isitobservable]

Thanks

[@tshepo2japhta28]

❤Thanks man

[@isitobservable]

Welcome 😊. Let me know if you should cover more topics related to grafana .

[@Faqarina]

the random explosion lol

[@Chrizz604]

Is it not possible to run "rate" on the response of a "delta"? We collect all the metrics in a summery value so I do have to use "delta" so I can see the change of the value, however I don't really know how I can run multiple functions on it.. e.g "rate(delta(api_dispatches{}[25s][1m]))" does not return anything

[@isitobservable]

Hi Chrizz, Very good questions. I could use several functions but in your case rate( delta i'm not sure it would work. rate only works with counters. delta is not returning counters. i guess you would need to do delta ( rate (...) instead.

[@ziaurrehman4738]

Did you create blog post to capture all these concepts and examples?

[@isitobservable]

Hi , Here is the blog version on this episode : isitobservable.io/observability/prometheus/how-to-build-a-promql-prometheus-query-language

[@olivierboisse1678]

20:53 Why is it forbidden to label filter with an empty string ? I tried prometheus_http_requests_total{unknownLabel=""} and it seems to be working

[@isitobservable]

Hi Olivier, it was initially a recommendation from the community. It seems that it is now tolerated...but you will need to filter out your metrics having empty label , or set a default value on your exporter.

[@z1lv1n4s]

Love your video. One note though - check how gauge is pronounced ;)

[@yotammedini509]

Nice. It would be helpful to see how such labels can pushed via Python client calling prometheus_client.push_to_gateway(...)

[@isitobservable]

Hi, Are your trying the find the impact of using a large number of labels within a metric? If so we could try to design a "load test" doing this and measure the impact ( in terms of memory).

[@u.hemanthkumargowd826]

can you show me hot to do run prometheus query using terminal

[@isitobservable]

First of all thanks for the idea, i will try to create a short video on this. FYI there is a small project that will help you achieve this : github.com/nalbury/promql-cli

[@haralc]

The content I'm expecting on a video with a title starting with "How to ..." is with computer screen and step by step tutorial. Not a commentary style like those KZbinrs who cannot do benchmark on the new hardware they were sent.

[@isitobservable]

Hi , Thanks for your feedback. I'm always trying to provide tutorial with the episode. Let me know how i could improve the tutorial. Thx

[@haralc]

I don't understand PromQL. After watching this video, I still don't understand PromQL ...

[@isitobservable]

Sorry to hear that. what are you trying to achieve ?

[@garypaduana62]

I tried, but I just can't follow along with the accent. Closed captioning was getting a lot of it wrong too. Sorry

[@josephmbimbi]

Tbh, from your explanations alone, i had a hard time understanding the difference between instant vector, range vector, the rate function. As there are lots of positive comments and review, i guess some ways of explaning are not suitable for everyone.

[@isitobservable]

Thanks for your feedback. i will work soon on the blog version of the episodes. so i will try to improve the explanation.