The problem with using avg_over_time for this is that it will ignore
missing data points.
If I understand Jennifer's request correctly, she's looking for a way to
start with data like {timestamp, value}, say the following values:
{0, 1}
{10, 1}
{20, 0}
{40, 1}
{50, 1}
and consider the missing {30, x} as a 0.
so in that example, avg_over_time would return (1+1+0+1+1)/5 = 80% and
Jennifer wants (1+1+0+0+1+1)/6 = 67%
My understanding of the original question is how to obtain the 6.
The numerator is easy: sum_over_time.
For the denominator the best I can come up with is:
count_over_time((vector(1))[5m:]), which is a really weird way of asking
"how many samples should there be in this range?"
Marcelo
On Wed, Jan 13, 2021 at 9:19 AM Ben Kochie <[email protected]> wrote:
> If you're looking for a "percent of OK scrapes" you can use:
>
> avg_over_time(probe_success[$__range])
>
> Then in Grafana, you can select "Percent (0 - 1)" to display the ratio as
> a percent.
>
> probe_success will always be returned by the blackbox_exporter is
> functioning properly.
>
> On Tue, Jan 12, 2021 at 5:26 PM Jennifer K <[email protected]>
> wrote:
>
>> Basically, I'm trying to fix an error that I am having when calculating
>> the percent. What i'm seeing:
>> sometimes the number of "good" scrapes is higher/smaller than the number
>> of scrapes in [$__range]. I think this is just because sometimes I get one
>> extra scrape compared to the number [$__range] scrapes and sometimes it's
>> spot on. I can't have a dashboard that sometimes shows the right value.
>> Therefore, trying to make this work and not show values over/under 100
>> when the value should be 100%. I am trying to "fix" the denominator to be a
>> value from a rule instead of the [$__range].
>> to do this I know I have "successful/unsuccessful probes" and "absent
>> data". I can quantify the probes using "probe_success" but when adding the
>> absent function to "probe_success" it doesn't work- I think it has
>> something to do with the vectors being different which is why I was trying
>> to use the ignore function.
>>
>> so... to answer your question-
>> I would want to count scrapes over the specified interval to include
>> absent data as well- basically if my scrape is set to 30s, an interval of 1
>> hour should return 120 (to include successful/unsuccessful and absent data)
>> what is the best algorithm to use for that?
>> Thanks so much!
>> Jennifer
>>
>> On Tue, Jan 12, 2021 at 9:40 AM Julius Volz <[email protected]>
>> wrote:
>>
>>> Do you want to count across scraped instances at *one* point in time, or
>>> do you want to count scrapes of a / each single instance *over* time?
>>>
>>> On Tue, Jan 12, 2021 at 1:10 AM Jennifer K <[email protected]>
>>> wrote:
>>>
>>>> to anyone that can help...
>>>> I've been trying to a total number of "scrapes" by adding successful,
>>>> unsuccessful and absent points together
>>>> this algorithm isn't working- any anyone explain why-
>>>>
>>>> (probe_success==bool 0) + (probe_success == bool 1) + ignoring (target)
>>>> sum without (target) (absent (probe_success))
>>>>
>>>> should be simple, but I just can't get it to work- any help would
>>>> greatly be appreciated.
>>>> Thanks!
>>>> Jennifer
>>>>
>>>> --
>>>> 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 [email protected].
>>>> To view this discussion on the web visit
>>>> https://groups.google.com/d/msgid/prometheus-users/a8b8a4a0-59e7-4087-bba6-a54252ebafdcn%40googlegroups.com
>>>> <https://groups.google.com/d/msgid/prometheus-users/a8b8a4a0-59e7-4087-bba6-a54252ebafdcn%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>> .
>>>>
>>>
>>>
>>> --
>>> Julius Volz
>>> PromLabs - promlabs.com
>>>
>> --
>> 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 [email protected].
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/prometheus-users/CAJ2Bn3q9xEADOb0r_nxOx4fXbGBYRo3tU%2B0nE8AtWD%2BqoiPCzw%40mail.gmail.com
>> <https://groups.google.com/d/msgid/prometheus-users/CAJ2Bn3q9xEADOb0r_nxOx4fXbGBYRo3tU%2B0nE8AtWD%2BqoiPCzw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
> --
> 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 [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/prometheus-users/CABbyFmqyphKcJS2eK10CQkzz8t3D3t3LDtbyo%3D4%2BuomygXNC8A%40mail.gmail.com
> <https://groups.google.com/d/msgid/prometheus-users/CABbyFmqyphKcJS2eK10CQkzz8t3D3t3LDtbyo%3D4%2BuomygXNC8A%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>
--
Marcelo Magallón
--
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 [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/prometheus-users/CABiJYgY%2BcboQcV2r8z%3DueL7pL69X5YiXYmJ2D3KQtaQ2QJ61vg%40mail.gmail.com.
