Hello, My apologies for the late reply.
No the data isn't getting deleted faster than the consumer is reading it. We also notice when this happens the number of Replica Fetchers doubles on the broker side and the total of out bytes drops significantly. Regards, Amine On 2019/09/07 04:18:18, Hrishikesh Mishra <sd.hri...@gmail.com> wrote: > Can you check whether its happening because logs are getting purge very > fast. > > On Sat, 7 Sep 2019 at 12:18 AM, Aminouvic <amine...@gmail.com> wrote: > > > Hello all, > > > > We're noticing several logs on our consumer apps similar to the following : > > > > 2019-09-06 17:56:36,933 DEBUG > > org.apache.kafka.clients.consumer.internals.Fetcher - Ignoring fetched > > records for mytopic-7 at offset 45704161910 since the current position is > > 45704162370 > > > > Any idea on what would cause this ? > > Did anyone ever encounter this kind of logs? > > > > Kafka Cluster version 0.10.1 > > Kafka Client version 0.10.2 > > > > Best regards, > > Amine > > > > > > >
