You may be hitting this scenario from my experience: As you have three nodes you begin to get deadlocks between load tasks.
These deadlocks cause tasks to be postponed, but real troubles happen when they survive past 30 seconds and are dumped to logs. There's massive amount of data in your tasks and they are pretty-printed for logging, leading to massive spikes in RAM usage and eventually OOM. The solution here to use smaller batches. How much data do you pass with one task currently? Regards, -- Sent from: http://apache-ignite-users.70518.x6.nabble.com/