Hi everyone - I'm hoping someone can shed some light on a stumper for me.
I've currently been migrating our log data from an ES 1.4 cluster to a shiny new 2.3 cluster. I'm pushing in quite a bit of Logstash data (150M+ documents per day) and I've noticed a huge discrepancy between the disk usage per index between 1.4 and 2.3. A daily index was ~200G on 1.4 and is over 450G on 2.x. I've made sure my templates are the same as previous versions and the mappings look fairly similar.
I've searched and searched but can't find any particular information on this - is there any known reason stores would be significantly larger on 2.x vs 1.x?
Thanks so much!