Recently encountered an issue with Elastic beanstalk where the application started giving error "host not found or unreachable" while connecting to Redis cluster. First I thought it might be related to AWS scheduled maintenance but maintenance was not started yet. On investigating further, I tested that the Redis cluster was live and was connecting from other instances.
On google, there was not much help available except one on amazon forum that amazon does restricts DNS requests after some limit and by default no DNS caching service is installed in Amazon AMIs.
After installing nscd, the issue was solved. I used the following .ebextensions/install-nscd.config to always install nscd and keep it running on any new instance added to the scaling group.
# these commands run before the application and web server are
# set up and the application version file is extracted.
packages:
yum:
nscd: []
services:
sysvinit:
nscd:
enabled: true
ensureRunning: true
Lesson learnt, always install DNS caching service on amazon AMIs.