When the EdgeView is started on the EVE device, the controller may install a script for client to run EdgeView commands on user's laptop. The script may have the filename of 'run.{device-name}.1659561808.edgeview.sh' which 'device-name' is the string for the device enabled with EdgeView session, and the digits are the timestamp in UNIX seconds when the EdgeView started.
The user can locate the script and run for example:
./run.{device-name}.1659561808.edgeview.sh -inst 1 route
On a Linux host, the EdgeView client script can be:
docker run -it --rm -e EDGEVIEW_CLIENT=1 -h=$(hostname) --network=host -v /tmp/download:/download lfedge/eve-edgeview -token <jwt-token> "$@"
On a Mac laptop, it does not support docker run in network in host mode, the port mapping for TCP can be explicitly defined in the script:
docker run -it --rm -e EDGEVIEW_CLIENT=1 -h=$(hostname) -p 9001-9005:9001-9005 -v /tmp/download:/download lfedge/eve-edgeview -token <jwt-token> "$@"
Although in EdgeView multiple instances, a user may run two EdgeView sessions on the same laptop for instance 1 and instance 2. Using explicit range will collide for the ports on laptop. Here is a more general script to set the part mapping range base on user's instance number input:
for var in "$@"
do
if [ ! -z "$gotinst" ]; then
instid=$var
break
fi
if [ "$var" = "-inst" ]; then
gotinst=1
fi
done
portstart=9001
portend=9005
if [ ! -z "$instid" ]; then
if [ "$instid" -lt 1 ] || [ "$instid" -gt 5 ]; then
echo "inst needs to be in the range 1-5"
exit
fi
newid=$(($instid - 1))
multi=$(($newid * 5))
portstart=$((9001 + $multi))
portend=$((9005 + $multi))
fi
PORTMAP="-p $portstart-$portend:$portstart-$portend"
docker run -it --rm -e EDGEVIEW_CLIENT=1 -h=$(hostname) $PORTMAP -v /tmp/download:/download lfedge/eve-edgeview -token <jwt-token> "$@"