Preparedness Script - Report challenges and fix what we can

[CarlB-dev]

To ensure a successful deployment, we can query devices to see number of virtuals, inspect REST settings, inspect snmp settings and if it is driving MCPD usage.

Determine high MCPD CPU usage by SNMP Can be leveraged to understand increased base CPU load.

K000136905: CPU and Memory tips when automating with REST, Ansible, AppViewX, AS3 and others

[CarlB-dev]

# Troubleshooting high mcpd CPU
# src - https://my.f5.com/manage/s/article/K000139100
# Current values
cd /var/tmstat/blade
tmctl -q -f mcp_stat_segment mcp_transaction_stat -s user,total_transactions,process_time_mean | awk '{use=$2*$3; print $0"   "use}' | sort -k 4 -rn | head -3

# Top 3
tmctl -f mcp_stat_segment mcp_transaction_stat -s user,total_transactions,process_time_mean -OK total_transactions -L 3

# Looking at historical snapshot data
# Change directory to /shared/tmstat/snapshots/blade0/public/3600 (or 86400). On a multi-blade capable system replace blade0 with the primary blade number, often 1.
cd /shared/tmstat/snapshots/blade0/public/3600
# Pick a snapshot file and put in place of <blade0-public-3600-20XX-XX-XXTXX:00:00> placeholder below
tmctl -q -f  <blade0-public-3600-20XX-XX-XXTXX:00:00> mcp_transaction_stat -s user,total_transactions,process_time_mean -OK total_transactions | awk '{use=$2*$3; print $0"   "use}' | sort -k 4 -rn | head -3
# You can also see how this is increasing over time in the snapshot files:
# Note: In this example we are looking at SNMP transactions over time. You can use this for any other process. You may need to change user=snmpd, to user=%snmpd
# You are still executing this in the snapshot directory previously selected.
tmctl -D . mcp_transaction_stat -s time,user,total_transactions,process_time_mean user=snmpd | awk '/^2/{p3 ? delta=($3-p3)*$4 : delta=""; p3=$3; print $0"   "delta;next}{print $0}'

[mikempw]

Adding to roadmap