stepanresl.bsky.social
Lead technical consultant @ DataBrothers
Data Platform #MicrosoftMVP ~ #PowerBI & #MicrosoftFabric
š Links:
Blog: www.datameerkat.com
GitHub: https://github.com/tirnovar
203 posts
492 followers
513 following
Regular Contributor
Active Commenter
comment in response to
post
That is just one water drop in the Monday oceanā¦ š¤£
comment in response to
post
Thanks to @sessionize.com for this easy widget!
comment in response to
post
comment in response to
post
Can someone please also fix a year bug that makes the "expand" parameter in the ODATA Feed in ALL pipelines unusable?
comment in response to
post
If you send 100 request for data sources. It will stop answering. And start to falling into Error 500. I have even implemented 0.5s delay between that should be required based on documentationā¦ didnāt helped. I was able to recreate it in currently 6 tenants. Two months ago this workedā¦
comment in response to
post
@itsnotaboutthecell.com, @jaypowerbi.bsky.social O:)
comment in response to
post
I will dig deeper this weekend. I have intuition that I can, by a few tricks, get at least a table name (now I have just a name for the column) as a viewer, and if I do, this might be an interesting issue.
BTW this was in a report published on the web.
comment in response to
post
@nickyvv.bsky.social ha! Finally, I was able to upload it.
comment in response to
post
comment in response to
post
Yep... (second was just rhetorical question)
comment in response to
post
comment in response to
post
Hmm, I will add a bit to this. On the axis is the name of the month in Czech. After rendering in HTML, visual renderers contain the name of the field from the model in its attributes. This doesn't happen for all visuals, but it's all aspects. But for the X-axis of the column chart, it does.
comment in response to
post
4,252ms, including two renderings of the display() function, is all you need to get the table from #Dataverse to #MicrosoftFabric in Parquet format. This approach also auto-selects columns created by the user and renames them without prefixes. Main benefit? Current data! Every time.
comment in response to
post
But I still need to check if Dataverse truly doesn't have a delay on this endpoint.
comment in response to
post
Sometimes, I need the current state of data, and if I can't trust that I have them because of some integration scheduler for data exportsā¦ it seems easier for me to use the TDS endpoint to get these data when I need them š
comment in response to
post
Or, of courseā¦ importing data to Dataverse tables by API with notebooks.
In the endā¦ all of these integrations costs CU(s) of Fabric
comment in response to
post
Currently, my tests usually last between 10 and 16 minutes. Regarding real-time reporting, I would not consider Dataverse as a storage.
Also, this is not only their integration. The fabric table as a Virtual table in Dataverse is also somewhat interesting.
comment in response to
post
If you are okay with up to 60 minutes of delay
comment in response to
post
I usually take logs from the browser š but that KQL might be handy.
comment in response to
post
Not really. I can make it work. I am just pointing to UX issues that I am seeing.
comment in response to
post
This quick test in #MicrosoftFabric showed me a lot of interesting behaviors:
- There are rendering differences between Python and PySpark notebooks.
- You can use "temp" storage allocated to the notebook for many crazy ideas, even for rendering videos š (if you then store it into Lakehouse)
comment in response to
post
Idea to #MicrosoftFabric
comment in response to
post
I would like to know how the tables are fetched. I would also like to see these errors included in the official documentation, especially since the link from the error ID takes you there. Can we get a preview window or link to the mentioned output? (just to decrease the amount of required clicks)
comment in response to
post
š¤£ someone had to try it!
comment in response to
post
Just need to add movements š¤£š¤£
comment in response to
post
It is not running perfectly, but it gets better.
comment in response to
post
OMG YES!!! Doom in Fabric Notebook
comment in response to
post
I had to try something crazy... and the funny is that it was working for a moment XDD