<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: Large dataset options for reporting server in Report Server</title>
    <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1968078#M20195</link>
    <description>&lt;P&gt;There is not really a good solution for this on Report Server alone. What we do with models of this size is to build them as SSAS tabular models. Then we can use deployment tools and partitions for incremental processing and we just create pbix reports in live connect mode to SSAS. If you have SSAS on a separate server (which is a good idea to spread the load) you just need to ensure you have Kerberos setup.&lt;/P&gt;</description>
    <pubDate>Tue, 20 Jul 2021 09:50:43 GMT</pubDate>
    <dc:creator>d_gosbell</dc:creator>
    <dc:date>2021-07-20T09:50:43Z</dc:date>
    <item>
      <title>Large dataset options for reporting server</title>
      <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1967687#M20190</link>
      <description>&lt;P&gt;I'm looking at building a large report (&amp;gt;100M rows) and I can't seem to find a good workign strategy on report server.&amp;nbsp; On premium this is easy using deployment pipelines, or incremental refresh, etc.&amp;nbsp; But what options are there for report server? I.e. how do I have a small working dataset locally (desktop) and the full dataset on the report server?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Tue, 20 Jul 2021 07:49:54 GMT</pubDate>
      <guid>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1967687#M20190</guid>
      <dc:creator>gatapia</dc:creator>
      <dc:date>2021-07-20T07:49:54Z</dc:date>
    </item>
    <item>
      <title>Re: Large dataset options for reporting server</title>
      <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1968078#M20195</link>
      <description>&lt;P&gt;There is not really a good solution for this on Report Server alone. What we do with models of this size is to build them as SSAS tabular models. Then we can use deployment tools and partitions for incremental processing and we just create pbix reports in live connect mode to SSAS. If you have SSAS on a separate server (which is a good idea to spread the load) you just need to ensure you have Kerberos setup.&lt;/P&gt;</description>
      <pubDate>Tue, 20 Jul 2021 09:50:43 GMT</pubDate>
      <guid>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1968078#M20195</guid>
      <dc:creator>d_gosbell</dc:creator>
      <dc:date>2021-07-20T09:50:43Z</dc:date>
    </item>
    <item>
      <title>Re: Large dataset options for reporting server</title>
      <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969100#M20203</link>
      <description>&lt;P&gt;Wow, what a terrible workaround.&amp;nbsp; &amp;nbsp;But thanks for the confirmation.&lt;/P&gt;</description>
      <pubDate>Tue, 20 Jul 2021 20:09:10 GMT</pubDate>
      <guid>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969100#M20203</guid>
      <dc:creator>gatapia</dc:creator>
      <dc:date>2021-07-20T20:09:10Z</dc:date>
    </item>
    <item>
      <title>Re: Large dataset options for reporting server</title>
      <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969141#M20204</link>
      <description>&lt;P&gt;I managed to find a slightly better workaround (still terrible though) which worked in my scenario.&amp;nbsp; I am using explicit SQL queries to get my data, i.e. not just pointing to a table, but actual &lt;EM&gt;select .... from ...&lt;/EM&gt;.&amp;nbsp; I do the subset for development using a where clause.&amp;nbsp;&amp;nbsp;&lt;BR /&gt;&lt;BR /&gt;What I needed to do in PROD was to ignore the where clause filtering the data.&amp;nbsp; So what I did was created a service account to refresh the data on report server, say account was &lt;EM&gt;domain\svc_data_refresh&lt;/EM&gt;.&amp;nbsp; So then I just changed my query to:&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;LI-CODE lang="markup"&gt;select ... from table
where ...
and ((select CURRENT_USER) = 'domain\svc_data_refresh' 
  OR /* filter clause */ date_created &amp;gt; '2021-01-01')&lt;/LI-CODE&gt;&lt;P&gt;&lt;BR /&gt;This ignored the filter clause if the user is the service account.&amp;nbsp; This way when I refresh data in PBI desktop (using my account) it filters data.&amp;nbsp; The server using the service account refreshes the full dataset.&lt;/P&gt;</description>
      <pubDate>Tue, 20 Jul 2021 21:45:21 GMT</pubDate>
      <guid>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969141#M20204</guid>
      <dc:creator>gatapia</dc:creator>
      <dc:date>2021-07-20T21:45:21Z</dc:date>
    </item>
    <item>
      <title>Re: Large dataset options for reporting server</title>
      <link>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969283#M20205</link>
      <description>&lt;BLOCKQUOTE&gt;&lt;HR /&gt;&lt;a href="javascript:void(0)" data-lia-user-mentions="" data-lia-user-uid="9236" data-lia-user-login="gatapia" class="lia-mention lia-mention-user"&gt;gatapia&lt;/a&gt;&amp;nbsp;wrote:&lt;BR /&gt;
&lt;P&gt;Wow, what a terrible workaround.&amp;nbsp; &amp;nbsp;But thanks for the confirmation.&lt;/P&gt;
&lt;HR /&gt;&lt;/BLOCKQUOTE&gt;
&lt;P&gt;It depends on what your use case is. In our scenario some of our models have 20+ reports linked to them using a live connection and we can then update a measures in one place and any fixes/improvements flow through to all those reports. And adding a page or changing a property on visual on a report does not require re-processing the entire dataset - it can be deployed instantly.&lt;/P&gt;</description>
      <pubDate>Tue, 20 Jul 2021 23:39:28 GMT</pubDate>
      <guid>https://community.fabric.microsoft.com/t5/Report-Server/Large-dataset-options-for-reporting-server/m-p/1969283#M20205</guid>
      <dc:creator>d_gosbell</dc:creator>
      <dc:date>2021-07-20T23:39:28Z</dc:date>
    </item>
  </channel>
</rss>

