Is the issue already resolved where when cgminer is restarted that the correct pool (the one of the customer) is connected to again? And also his/her backup pools?
Enable save command and will work.
FYI, for those using BAMT, the save call in the api will create a new cgminer.conf under /.cgminer/cgminer.conf .
This is of course useless since BAMT starts cgminer pointed to /etc/bamt/cgminer.conf
A workaround is to schedule a job to check for updates to /.cgminer/cgminer.conf and overwrite /etc/bamt/cgminer.conf . This will restore your customers pools in the event that you need to reboot/power outage/etc. Unfortunately the cgminer.conf created is a bit over-complicated, and duplicates values for each of your GPUs, even if the values are the same for all of them... regardless it does work, and is the best for the customer.
Login to your BAMT box using ssh and perform the following:
sudo crontab -e
Add the following to the bottom of the crontab (This will check for updates every 5 minutes and only overwrite when changes are made)
*/5 * * * * cp -u /.cgminer/cgminer.conf /etc/bamt/cgminer.conf >/dev/null 2>&1
Exit (and save)
Also, please keep in mind that SMOS linux has a script that will periodically mine for the developers, thus robbing your customers of time. Use BAMT or SMOS bee edition, or manually remove the offending code.