Forum Discussion
Helper IVIs there any maximum size limit and data storage limit when using Power BI Report Server?
Hey Anonymous
Yes we use SSAS Tab for those very reasons, scale, RLS and a managed data model approach.
The data query chat thing is a killer. We have a distrib SSAS solutions 4 servers across 2 data centres behind an L5 alias. Then a pile of PBI Servers up front again under and L5 alias. Its way, way less responsive than the much smaller dev environment whihc has PBI-SSRS and SSAS TAB on the same box. the L5 is getting absolutely battered when we spin up any significant PBI load.
Maps will just add to the problem, we're lucky as we don't do very little meaningful geographic stuff.
We are hoping for multiple SSAS Tab models too! Though we have scenarios that involve imported Excel linking to live SQL or SSAS Tab linking to Live SQL that are going to blow away a ton of nasty hacks/workarounds we have.
Be warned my understanding is based on observation and some informed guesswork rather than anything concrete so don't take it as gospel.
Thanks for the explanation!
My case is as follows:
- I have ElasticSearch cluster with weekly indices.
- Each index has 200GB of data.
- Power BI obviously can't query ElasticSearch directly.
But, I managed to query data from Power BI Desktop in different ways:- DAX (Custom Qury), with Web.Contents
- ASP.NET Service (Web)
- CData ElasitcSearch ODBC.
We also have CData Power BI Connector, which is a custom conntector. Unfortentlly, DirectQuery with Custom Connectors still supported only with Power BI Desktop (nieghter in Power BI Server / Power BI Server, as far as I can see). Link here.
- I'm trying to query 4 indices, which means basically 800GB. I'm trying to query only a few properties from each document, not all the document. It is very slow to query many documents, as each request only get 10,000 documents at once. However, caching it though Power BI data-set which refreshes every day is fine - if it works (haven't tried it yet). I'm affraid this data set would be too big, so maybe i'll store a seperate data set for each weekly index and merge them later.
- How complex is to import the data to some SQL Server / Service Analaysis? Remember it's JSON documents and some of them are very complex.
{
"ProductId":"88432",
"Comments":[
{
"Username":"stpnet",
"Comment":"Cool Product!",
"Stars":5
},
{
"Username":"rblatchford",
"Comment":"Excellent!",
"Stars":4
}
],
"Tags":[
{
"Key":"Category",
"Value":"Electricity"
},
{
"Key":"IsNew",
"Value":true
},
{
"Key":"ShippingType",
"Value":"Worldwide"
}
]
}