lazyant 12 hours ago | parent | flag | favorite | on: I usually run 'w' first when troubleshooting unkno...

My goto for initial troubleshooting a server is:

uptime # uptime and CPU stress

w # or better yet:last |head # who is/has been in

netstat -tlpn # find server role

df -h # out of disk space?

grep kill /var/log/messages # out of memory?

ps auxf # what's running

htop # stressed? , look out for D (waiting on I/O typically) processes

history # what has changed recently

tail /var/log/application.log # anything interesting logged?