Which is more efficient? KV, Dissect, Split, Grok

Elastic Stack Logstash
#1

Hi all,

As there are many ways to achive similar goal using logstash filters, would like to discuss and compare between KV, Dissect, Split, and Grok, which is a better way of handling data?

Scenario 1:
Mapping & Parsing logs with consistent delimiter

Better: KV

Scenario 2:
Mapping & Parsing logs with inconsistent pattern

Better: If both Dissect and Grok able to achieve the same goal, should we use Dissect or Grok for a more efficient / effective parsing?

Scenario 3:
Extracting data from field (For example extracting server name fqdn, or domain from a url)

Better: If both Split / Grok able to achive the goal, which will be more efficient to use?

Anyone have any example would be nice to discuss together too!!

Thanks!!

#2

In general I would say that the extreme flexibility of grok comes at a price compared to dissect.

That said, unless you are processing very large volumes of data it does not make sense to choose a filter based on cost. Instead choose whichever one is simpler. If you are processing huge volumes of data, to the point where cost is a significant factor, then benchmark each option and measure the cost.

1 Like
#3

Thanks for the great tips Badger!! It make sense that if cost isn't a big factor we should choose whichever simpler!!

Kudos!