Heartbeat schedule not work when has large number of hosts

Elastic Stack Beats
1

heartbeat icmp monitor not work with large number of hosts with config schedule: '1 * * * * * *'
It did not show any error but cannot create index at elasticsearch .

not work (under 4096 hosts it is normal)
schedule: '1 * * * * * *'

work (more than 30000 hosts)
schedule: '@every 60s'

ps operating systems centos 8 already set file descriptor limit to 65535 . all hosts at same yml

2

Hmmm, this seems related to https://github.com/elastic/beats/pull/22369 . I haven't been able to replicate it myself, but I only tried with a few hundred pings. I'll try later this week with 10k or so.