AWS Vertica timeout exceptions
Hi,
I have a problem on my AWS Vertica (7.1.1-10) cluster site on CentOS 6.
I did a copy cluster from Vertica (7.1.1-10) to AWS Vertica site (CentOS-6 too)… and now I'm getting timeout exceptions only on AWS site (thing that doesn't happening on the source cluster)
I'm connecting to Vertica with DBeaver, DataGrip and Jenkins… and in all programs I'm getting the following error after several minutes:
[HY000][100024] [Vertica][VJDBC](100024) IOException while communicating with server: null. Refer to the inner exception for more detail. java.io.EOFException
I tried to set some parameters (set them to be same as the source cluster):
# cat /proc/sys/net/ipv4/tcp_keepalive_time
120
# cat /proc/sys/net/ipv4/tcp_keepalive_intvl
120
# cat /proc/sys/net/ipv4/tcp_keepalive_probes
45
But it didn't helped… Still having the problem 
Do you know what can cause that?
How on one environment I'm not getting this error and in AWS environment I do?
Comments
Fixed
We are having DNS Cname on AWS Vertica cluster,
We increased idle timeout to 5 min from 30 sec
Thanks
Chen