This is a good question but can get a tad confusing. So I hope I can shed some light on it without making more confusion/questions.
So in your case, we have two hosts, Host 1 = 16Gib and Host 2 = 32Gib. Both of them running at different times resulting in non-concurrent timing. which can be depicted as such.
So what I would expect you to pay, per your rate card.
- 1 Hour for Host 1, because the monitoring time incurred is 30mins which be rounded to the nearest hour when usage is greater than 5 mins.
- 1 Hour for Host 2, because the Size of the host doubled, So Dynatrace isn't seeing this as a 1 to 1 continuation of the powered down host. This too has a 30 min monitoring window where the usage is greater than 5 min, so it gets rounded to the nearest hour.
Here is the verbiage on the 5 min monitor rounded: "If a host runs for less than 5 minutes, it doesn't count against your host unit hour quota. A host running for 5 minutes or longer is rounded up to 1 host unit hour."
So, all in all, Host 1 = 1HU x 1 Hour x $<Rate Card Value> = Host 1 Cost | and | Host 2 = 2HU x 1 Hour x $<Rate Card Value> = Host 2 Cost | Add both resulting values = your total cost to monitor.
I have put in a request for Dynatrace to out line the Hours monitored for hosts, That way if a host spins up and spins down during peak hours or business hours, it would be nice to look at a host over a 6 month spread and say "Ahh, this has only been monitored for 730 hours which = 1 Month. Therefore you'd be able to really formulate showback/billback to charge teams exactly for what they use per host.
So in short, Both Hosts will incur a 1 hour monitoring segment resulting in a value of 2 total hours (Summed) then you'd add in the Host Unit Multiplier where Host 1 = 1 HU and Host 2 = 2HU.
Thank you for the answer Chad, you explain it excellently!
I want to confirm the HUH usage for the case above. If we ignore the Rate Card Value, the total of HUH usage is (1HU x 1 Hour) + (2HU x 1 Hour) = 3 HUH isn't it?
And we have one more question. The case is, there are 3 hosts (2 hosts 16GiB and 1 host 32 GiB) running serially like illustrated in picture below.
I assume the calculation will be like this:
The cost of Host 1 is 1 HU, then it is spun up with Host 2 (which also cost 1 HU), both will cost 1 HUH. After that, the Host 3 costs 2 HU x 1 Hour = 2 HUH.
So, the total will be additional of 1 HUH (from Host 1 and Host 2) with 2 HUH (from Host 3) and the result is 3 HUH. Please kindly correct me if my calculation for this question is wrong.
And again, thank you!
You are close to 100% correct and your time table is so much nicer than mine!
So lets just say your Full stack costs $60 for Each HUH. Host 1 and 2 can be viewed as a continuation of cost/monitoring from the 15 mins on Host 1, to the 15 Mins on Host 2 which will total you to 30 Mins of monitoring that is then rounded UP to 1 hour. Ultimately costing you 1 Hour X 1 HU X $60 = $60.
Bringing in Host 3 which is doubled the size, using the same rate card value. It has a 30 Min monitoring duration but is larger than the Host 1 and 2 so it is not seen as a continuation of monitoring. So Host 3, is monitored for 30 Mins which is rounded UP to 1 hour. Ultimately costing you 1 Hour X 2HU X $60 = $120.
Now we can SUM up the two costs for the overall 3 hosts, being $180 which covers monitoring from the 10-11 time slot with the 3 hosts that were spun up.
You have two errors
1 - "The cost of Host 1 is 1 HU, then it is spun up with Host 2 (which also cost 1 HU), both will cost 1 HUH" - These both will not bill you independently as a 1 resulting in 2 HU, rather, it is a continuation of 1 host at the HU cost Rounded to the nearest full hour.
Example 1
You have a host with 16 GiB RAM (which equals 1 host unit) running from 10:00-10:30 AM. At 10:30 you spin up another host of the same size. Dynatrace considers this a single host unit because the hosts don't run concurrently.
Example 2
You start the first host at 10:00 AM and launch another host at 10:30 AM. Then, both hosts run together for 30 minutes and are shut down at the same time. Dynatrace considers this to be 2 host units because both hosts run at the same time.
2 - "So, the total will be additional of 1 HUH (from Host 1 and Host 2) with 2 HUH (from Host 3) and the result is 3 HUH. Please kindly correct me if my calculation for this question is wrong." - The Host Unit Hours (HUH) incurred is 2, because Host 1 and Host 2 are rounded UP to the nearest 1 Hour as Host 2 is just a continuation of Host 1 Monitoring, While Host 3, being its doubled size, is viewed as net new, and will be rounded UP to 1 Hour, resulting your HUH to be 2.
When we SUM up the HU (Host Units) Both Host 1 and 2 are conjoined as a continuation, resulting in a cost of 1 HU. Host 3 being larger, starts its consumption at a HU of 2. Resulting on the overall SUM, for the 3 hosts, as 3 HU's.
But each one of these hosts, with their associated HU's and HUH's should be formulated as a line item cost. (This works for us since our hosts are up 24/7) I do this in excel where I list the Host name, the DT ID, The monitoring Mode, and the RAM size installed. From there I have equations/formulas that bucket the RAM to determine the HU, then I take that HU value and assume its a 365 Days of Monitoring, which then formulated the assumed maximum cost of monitoring for a year given. Now that's a high water mark, If the host is offline during non business hours then that cost would be drastically reduced. To which my hope is that Dynatrace adds in the Monitoring Hours as metadata on the Gen 3 page of hosts. Or something that I can pull via the API.
Featured Posts
