I have a requirement of showing data tables having approx 8 to 10 columns with large data (approx 250k or more).
But it seems like fetching of data is successful but Kibana cannot render the table and webpage become unresponsive.
I want to know is it a known problem of Kibana and what can be the possible solution for this?
Would you be able to share a screenshot of your data table configuration in the visualization editor? I wonder if it could have anything to do with formatting the type of metrics or aggregations.
Hi Tim,
I am having exactly the same issue mentioned by @jay_shah. I have got 50K documents in total. I am using 6.1.4 (elk). I have data based on years starting 1979 till date. Data loads correctly if the time range is for 30 days, 60 days or a 1 year. After that the visualization is too slow to load if I go beyond that. After looking at each visualization carefully data table visualization appears to be the bottleneck the Bucket Size is field if it increases the visualization become slow. In general I have 52000 unique records, am not show on what basis should I set this in the data table visualization. Any help would be great !
6.2.x and 6.1.x are the previous minor version from the latest, which is 6.3.2. Not sure if it would be possible for either of you, but I'm curious if a staging environment of 6.3.x could be set up and we can see if the problem still happens with the latest version.
Hi @tsullivan,
thanks for standing by
please have a look at this temporary vm 145.100.58.19:5601 and look at rasff_Dashboard_V3. If you change the date for 30 days the performance is ok, but more than a year the browser crashes the problem is with data table Bucket size field. Can you please have a look at it ?
Background of the VM I have got 200 GB main memory and I have set the heap size of elasticsearch to 32gb in the jvm.options file in the config directory I have made the following settings
-Xms32g
-Xmx32g
Thanks for the temporary VM with the data, that does help!
I looked at the query generated by the table visualization, and ran it in Dev Tools. It isn't taking Elasticsearch a long time to generate the result, but the result is a HUGE amount of JSON data. I copied that result data into an editor and it's 416,757 lines.
When I went into the visualization editor (don't worry, I didn't change anything) to run that visualization by itself, it took awhile for Elasticsearch to return the response payload, but not terribly long.
This definitely looks like a front-end performance issue, as Elasticsearch is actually performing very well for the search (that 32-gig heap size certainly helps!)
I captured a performance profile of the work that the browser is doing while this table renders, and I'm going to share it with the Kibana visualization team. They're already aware of some performance issues in this visualization, but I don't think there are any issues filed on it. I have filed https://github.com/elastic/kibana/issues/21739, which we can follow for updates from the engineering team.
Alternative approach:
Please have a look at rasff_Dashboard_V4 at the same url . Here instead of using "data table" visualisation type I am using a saved search called table that I created in Discover. Now embedding this within this dashboard the performance here appears to be perfect !
However I am not able to provide my users possibility to download these data as a CSV format. Is there a quick hack that I can utilize to provide this functionality ? or alternative suggestions ?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.