PowerScale: Isilon: OneFS: How to install and run internal_ping script to test back-end network connectivity

There is a single-line shell script, or run through most current OneFS health check script.

One line option:       

FILE=/ifs/data/Isilon_Support/backendpingtest.txt; echo "" > $FILE; isi_for_array -XI 'for i in $(isi_nodes %{name}":"%{internal_a_address}":"%{internal_b_address}":"%{failover}); do name=$(echo $i | cut -f 1 -d ":"); if [ "$name" != "$(hostname)" ]; then iba=$(echo $i | cut -f 2 -d":"); ibb=$(echo $i | cut -f 3 -d":"); ibf=$(echo $i | cut -f 4 -d":"); ping -c2 -t2 $iba > /dev/null; if [ $? -eq 0 ]; then iba_ok="|A| _OK_ "; else iba_ok="|A| FAIL "; fi; ping -c2 -t2 $ibb > /dev/null; if [ $? -eq 0 ]; then ibb_ok="|B| _OK_ "; else ibb_ok="|B| FAIL "; fi; ping -c2 -t2 $ibf > /dev/null; if [ $? -eq 0 ]; then ibf_ok="|F| _OK_ "; else ibf_ok="|F| FAIL "; fi; echo -e "-> :$name\t$iba_ok\t$ibb_ok\t$ibf_ok"; fi; done' >> $FILE && clear; cat $FILE | sort -k1,3 -V | grep -v "Host is down" | tee /ifs/data/Isilon_Support/backendpingtest-$(date +%F_%H%M%S).txt

Let the command run through. If you want to stop the command early, run CTRL-C. If it does not print anything, it does not mean it is not working.

This can be run anytime on the cluster and must be run on a single node.

Isilon-1: -> :Isilon-2   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-1: -> :Isilon-3   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-1: -> :Isilon-4   |A| _OK_  |B| _OK_  |F| _OK_
Isilon-2: -> :Isilon-1   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-2: -> :Isilon-3   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-2: -> :Isilon-4   |A| _OK_  |B| _OK_  |F| _OK_
Isilon-3: -> :Isilon-1   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-3: -> :Isilon-2   |A| _OK_  |B| _OK_  |F| _OK_ 
Isilon-3: -> :Isilon-4   |A| _OK_  |B| _OK_  |F| _OK_
Isilon-4: -> :Isilon-1   |A| _OK_  |B| _OK_  |F| _OK_
Isilon-4: -> :Isilon-2   |A| _OK_  |B| _OK_  |F| _OK_
Isilon-4: -> :Isilon-3   |A| _OK_  |B| _OK_  |F| _OK_