Re: commiting consumed offsets synchronously (every message)

2018-02-21 Thread Sönke Liebau
> Nicu > > From: Sönke Liebau [soenke.lie...@opencore.com.INVALID] > Sent: Wednesday, February 21, 2018 4:59 PM > To: users@kafka.apache.org > Subject: Re: commiting consumed offsets synchronously (every message) > > Hi Nicu, > > committin

RE: commiting consumed offsets synchronously (every message)

2018-02-21 Thread Marasoiu, Nicu
possibility of duplicate processing downstream of a topic. Thanks for your help, Nicu From: Sönke Liebau [soenke.lie...@opencore.com.INVALID] Sent: Wednesday, February 21, 2018 4:59 PM To: users@kafka.apache.org Subject: Re: commiting consumed offsets

Re: commiting consumed offsets synchronously (every message)

2018-02-21 Thread Sönke Liebau
Hi Nicu, committing after every message and thus retrieving them with a batch size of 1 will definitely make a huge difference in performance! I've rigged a quick (and totally non academic) test which came up with the following numbers: Batching consumer - Consumed 1000490 records in 5 seconds No

commiting consumed offsets synchronously (every message)

2018-02-21 Thread Marasoiu, Nicu
Hi, In order to obtain an exactly-once semantics, we are thinking of doing at-least-once processing, and then have a compensation mechanism to fix the results in few minutes by correcting them by substracting the effects of the duplicates. However, in order to do that, it seems that at least thi