Prometheus Query help

[Meher Garda]

Hi

We have a service setup in active/passive mode.

Each service spits out two metrics, lock_state and request_process_time

both these metrics have the following common labels: env (dev or qa or prod) , instance (instance ip) , stack (consul service name )

With query 

lock_state{stack=~"failoversvc.*"} == 1 , I can find out which instance has the lock and is active

With query

request_process_time{stack=~"failoversvc.*"} , I can get the value for this metric

Is it possible, to run the second query only for the instance which has the lock ?

[Brian Brazil]

request_process_time{stack=~"failoversvc.*"} and lock_state{stack=~"failoversvc.*"} == 1

--

Brian Brazil

www.robustperception.io

[Meher Garda]

Hi Brian

Thanx , i kinda got to the same query but im seeing something weird.

request_process_time{stack=~"failoversvc.*"} and lock_state{stack=~"failoversvc.*"} == 1        <------- this gives me no result

But

request_process_time{stack=~"failoversvc.*"} and on (stack,instance) lock_state{stack=~"failoversvc.*"} == 1    <------ gives me data

Can you help me understand what is keyword `on` doing or what is it adding to the mix?

Regards

[Brian Brazil]

On 14 March 2017 at 16:44, Meher Garda <meher...@gmail.com> wrote:

Hi Brian

Thanx , i kinda got to the same query but im seeing something weird.

request_process_time{stack=~"failoversvc.*"} and lock_state{stack=~"failoversvc.*"} == 1        <------- this gives me no result

But

request_process_time{stack=~"failoversvc.*"} and on (stack,instance) lock_state{stack=~"failoversvc.*"} == 1    <------ gives me data

Can you help me understand what is keyword `on` doing or what is it adding to the mix?

I'd guess that there's other labels in play than the ones you listed, and this is causing them not to be considered.

Brian

 

Regards

On Tuesday, March 14, 2017 at 9:35:41 AM UTC-7, Brian Brazil wrote:

On 14 March 2017 at 16:15, Meher Garda <meher...@gmail.com> wrote:

Hi

We have a service setup in active/passive mode.

Each service spits out two metrics, lock_state and request_process_time

both these metrics have the following common labels: env (dev or qa or prod) , instance (instance ip) , stack (consul service name )

With query 

lock_state{stack=~"failoversvc.*"} == 1 , I can find out which instance has the lock and is active

With query

request_process_time{stack=~"failoversvc.*"} , I can get the value for this metric

Is it possible, to run the second query only for the instance which has the lock ?

request_process_time{stack=~"failoversvc.*"} and lock_state{stack=~"failoversvc.*"} == 1

--

Brian Brazil

www.robustperception.io

--
You received this message because you are subscribed to the Google Groups "Prometheus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to prometheus-users+unsubscribe@googlegroups.com.
To post to this group, send email to prometheus-users@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/24986391-a2ec-4ba1-a6da-d1858d570019%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Brian Brazil

www.robustperception.io

[Meher Garda]

Yes, there are other labels too, Thanx once again.

To unsubscribe from this group and stop receiving emails from it, send an email to prometheus-use...@googlegroups.com.
To post to this group, send email to promethe...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/24986391-a2ec-4ba1-a6da-d1858d570019%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Brian Brazil

www.robustperception.io