In the kibana logs of the RUM events (APM transaction events) at the Elastic Cloud APM servers, we have got the raw user agent string like
Mozilla/5.0 (Linux; Android 7.1.1; SM-C900F) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.90 Mobile Safari/537.36
whose corresponding device name comes as
On simply googling SM-C900F, we get which device it is, in this case, it is Samsung Galaxy C9 Pro SM-C900F.
If need be, I can post other such raw user agent string which are being tagged as Generic Smarthphone instead of their actual device name
This seems to be a limitation of our User Agent processor. I've created this issue to request this improvement for the user agent processor.
Please feel free to comment on the issue or follow it for updates.
Hi @Hamidreza is there a way we can add our custom processing to the user agent before the document gets indexed?
It is possible to define an ingest pipeline with multiple processors. In the User Agent Processor documentation page, for example, the pipeline is defined with only one processor, but you can add more to the
processors field when you define the pipeline.
Please also make sure to configure the APM server to use the custom pipeline.
This topic was automatically closed 20 days after the last reply. New replies are no longer allowed.