My repeaters have a consistent "flashing" behavior, upon opening a dashboard, is there a reason why? I'm referring to both (a) the intial whiteout and reappearance of the whole repeater, and (b) the repeated reappearance of the lines

Hi Daniel again, may I ask, is this the only viz with this behaviour?

it happens to any repeaters

Can I confirm is also in the Test environment?

ya

I have impersonated using our Admin user. I will create a Test dashboard in order to isolate the Repeater insight and have a better look in the the dev-console.

yes; i am currently using it

In the Data tab, once you click in Edit, and wait for a bit to charge the

l_schema

, there, in the right side of the screen are around 10 warning about mapping/validation, I would recommend first to fix the mapping of the columns on your Redshift datasource.

i see

Hi Daniel, Thanks for your patience. I noticed you've made changes to the dashboard (and changed the workspace), specifically in the 'Brand Information Dashboard' where the flashing has stopped. I see you've removed filters from the 'Most Loved Brands' visualization, leaving only 'All Brands'. Did you make other changes that resolved the blinking? Regarding your performance question: LDM mapping errors, especially with your Redshift data, cause significant performance issues. GoodData repeatedly attempts to fix these errors, increasing query times and resource use, which could be the cause of the flashing you observed. I can see still are the warning, I strongly suggest fix these warning and then see and finally confirm the flashing is gone (or not).

got it, thank you! yes it's a time of active development, and indeed I'm having performance issues with the dashboard. I'll clear things up and see where i stand

Please, let us to know once you have an update. Have a nice weekend.

i see that all of the mapping warnings regard the following scenario: a table has COLUMN A and COLUMN B COLUMN B serves as a label for COLUMN A Also, COLUMN B is mapped as itself, COLUMN B both the A>>B label and the B column by itself are used in visualizations, so changing the mappings will break a lot. is there any straightforward solution besides rewiring my LDM manually?

Hi Daniel, could you tell me, from your LDM, what are the Columns A and B (you can use DM)?

DM'ing you

Hi Daniel, After performing a few tests in both the PROD and DEV environments, I can confirm what you shared via direct message: there is clearly a slowness in the PROD environment. I ran both tests for five minutes, and in DEV, all the visualizations appeared in less than 20 seconds. However, in PROD, I was unable to see the mentioned visualizations and encountered several errors in the developer console, all related to the connection to the data source. It seems that the connection is being canceled for some reason. I noticed that you have two different connection URLs in the settings of your data source. Are the data from both URLs identical? I assume they are. If that's the case, why not connect to the same data source? I will share more details via direct message. Have you check the integrity of the data for the url on PROD?

Thanks Mauricio. We keep our dev and prod on separate tracks all the way upstream to our serverless workgroups. The data & connection are the same back there, or at least they're supposed to be. There's no difference in the data that accounts for this. but I appreciate you confirming that everything is identical as far as gooddata goes.

Hi Daniel, if there is not difference in the data and connection as you said (but still seems you are not 100% sure). I think it would be worth to try using the same connection in PROD, or as I wrote be sure on the integrity of the data or some schema/query. Thus I would check, the integrity of the data for PROD. And, also as a test can you try to replicate EXACTLY the set of vizualisations from DEV to PROD, includiing the difference in filters?

Hi Mauricio, alas i've confirmed that everything is the same as far as I can detect; we don't have a solution for now besides just attaching the PROD dashboard to our DEV backend. We've seen unexplained problems with redshift serverless before

Hi Daniel, this is a follow up, I would like to know if somehow the issue was fixed, or the connection to redshift was improved. Please, then let us to know if you need further assistance.

there was nothing i could do to fix this, unfortunately. We simply wired our prod dashboards to our dev backend

I apologize for the inconvenience. If the connection settings are correct and the data is present (which I believe it is), there's not much more we can do, as it's beyond our scope. Since you have encountered similar issues with Redshift in the past, we recommend reaching out to their support team for further assistance. They will be better equipped to troubleshoot any potential issues on their end. In the meantime, if you have any other questions or concerns related to GoodData, please feel free to reach out to us by creating a new thread. Have a good evening Daniel 🙂

hey! yeah it's happening in edit mode, absolutely. in the LDM too

Hi Daniel, thank you very much for your help.. it seems we will need to open a ticket on your behalf. Could you please 'play around' and reproduce the issue when you see the slowness and possible connection issue, for example, when open dashboard and also when you try to Edit the LDM. Also, can you provide us then the window time to check our logs?. I will add you as 'Requester', probably then you can provide even more details in there. We appreciate your help to solve this Redshift inconvenience.