Hello!
You can decrease number of stripes but it will also hamper your
performance. Moreover maybe queue will just be redistributed.
Have you tried decreasing rebalance throughput?
https://apacheignite.readme.io/docs/rebalancing#rebalance-thread-pool-tuning
Regards,
--
Ilya Kasnacheev
пт, 4
Hi,
I have not been able to get the heap dump after oom (it is not opening from
mat or jhat) but I am repeating my experiment and just took the heap dump
when one of the ignite servers is at 2 gb out of 5 gb.
This is what the mat extract tells me
80 instances of
Hi
I am running Ignite data grid in persistent mode with following
configuration.
20 clients
3 ignite servers (S1,S2,S3) of Xmx 5G started with 2.6.0
I have set WAL and WALArchiving to the same path so that the WAL Archiving
is disabled.
For TcpCommunicationSpi
After 15-20 minutes of run (by
Hi
I am running Ignite data grid in persistent mode with following
configuration.
20 clients
3 ignite servers (S1,S2,S3) of Xmx 5G started with 2.6.0
I have set WAL and WALArchiving to the same path so that the WAL Archiving
is disabled.
For TcpCommunicationSpi
After 15-20 minutes of run (by