Thanks for the insights, Sorry i didnt make myself clearer. I put the current Pi-zero system onto the unreliable power supply (router that gets rebooted when wifi plays up) in order to replicate and reproduce the problem after it occurred the first time on the 2nd system. Not an ideal setup but it does replicate the problem rapidly
My headless Pi-Zero systems have been getting more reliable since i have reduced the random s/w installed on them to look see and have only been installing the key s/w onto the base system that is getting used. but even so they do still on occasion drop the wifi connection and refuse to pick it back up again requiering the occasional restart (gpio triggered shutdown/restart should perhaps be my next priority,)
I have no idea how node red handles its persistent data stores, going by the gui settings for the node.
Usingt the interval setting which i have set to 3600 (ihour)
Code: Select all
the minimum buffer flush interval in seconds.
I was working under the assumption that this would open the data store file every hour and write the data buffered (within the node red chart node to the the file and then close it.
whatever the mechanism is that indexes system file write/commits and queues them up something about this particular file must be relegating it to last position in that list for it to be consistently the only file regardless of file name that gets effected.
So I will look into issuing a sync command from within node red, im also considering a cron job to copy the file as long as its bigger than the last file copied/backed up.
Or look at using a dbs system to record the data ..