Hello! How many nodes do you have?
Maybe when you start killing nodes, remaining nodes start rebalancing data and they keep more data locally than fits in your memory. The recommendation stays the same - decrease your data regions' and heap size until it fits comfortably in the available memory to not trigger OOMKiller. Regards, -- Ilya Kasnacheev ср, 14 апр. 2021 г. в 07:39, rakshita04 <rakshita.chaudh...@siemens.com>: > Hi Ilya, > > we could find out the problem of the application stop. > We see "Buffer Overflow" error when application stops. > We commented out mCache.Put() in our code(basically not calling Put API to > write data to DB) and the restart did not happen. > Also this restart happens after a certain number of entries in DB. > Do you have any idea what can cause this buffer overflow while write in DB? > and is there anything we can do to avoid it? > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >