We're looking to host DataMiner in Azure but cloud-hosting costs are a big concern.
Beyond the two threads below, does anyone have any suggestions or experience to share on ways to optimize the system in order to keep the overall running costs low?
Hi Bruno,
First of all, if you want to host it on-prem, it won't be free as well. You can find lots of resources on the internet giving you the pros and cons of Azure vs on-prem. So, I'm not going to list all the advantages here, instead I'll just link to one of those webpages: Azure vs on-premise: Why you should choose the cloud | Doherty Associates
Secondly, when hosting a DMA in Azure, you need to make sure you put a reservation on your VM, ideally of 3 years. This drastically reduces the pay-as-you-go price you see on the website. And today you can still stop a reservation without any cost, so don't worry about that 3 year, you can still stop or change earlier.
Thirdly, I would also like to mention our upcoming DataMiner storage as a service. With this service, you no longer need to set up your own Cassandra and Elastic cluster and you can just leverage the DataMiner storage service. This also allows you to only pay for what you use and scale as you go. This is coming up in Q3 of 2023.
Bert
The bigger your system, the more compute and storage resources you’ll need, and this is independent if you host it on-prem or in the cloud. The advantage of cloud, is that you can more easily scale up and down, and you don’t have to plan this up front.
Now, limiting trending/retention/polling might indeed reduce the need for those compute and storage resources, but personally for me this is independent if you host it on-prem or in the cloud. You need to make reasonable decisions on what needs to be trended, how fast and how long you need all your data. And based on those decisions, this will indeed result in a hosting cost, but if you made the correct and reasonable choices, the cost will be reasonable as well. E.g. It doesn’t make sense to trend everything while you don’t need it, and it doesn’t make sense to trend nothing while you do need history on those key metrics. I would say, use common sense, and the hosting price will also be reasonable, on-prem or in the cloud.
Thanks, Bert. Regarding the specific use of DataMiner can you point out some of the things that I need to be aware of that may impact hosting prices? Do I need to look at things like limiting real-time/avg trending of parameters, data retention, polling speeds, etc?