VM experiencing very high disk read cause by randomly process cron -f, google-cloud-ops-agent

[visiniaga]

just happening yesterday,  without any reason as previously normal,

system running:
- ubuntu 20.04 with latest update linux-gcp 5.15.0.1055.63~20.04.1
- php 7.4.3-4ubuntu2.20
- google-cloud-ops-agent 2.46.1~ubuntu20.04
- google-guest-agent 20231004.02-0ubuntu1~20.04.3

checking for ops agent monitoring, showing on process tab, vm got very high disk read:
high process from:
- kswapd
- /usr/bin/google_guest_agent
- /opt/google-cloud-ops-agent/libexec/google_cloud_ops_agent_diagnostics -config /etc/google-cloud-ops-agent/config.yaml 

cpu usage is normal for 20-30% for google cloud ops agent, with resident memory also not normal using 2gb in virtual memory report, but disk read taking full that vm freeze

vm running nfsd, moodle tools (ghostscript, pdftoppm, unoconv, clamav-daemon) 

vm running nginx and php

checking from syslog with bash shopt, from the last time vm showing log before freezing
- /usr/sbin/CRON -f 
- moodle cron.php

as workaround now, disable any user cron temporarily 
pls any suggestion or advice