admin Another busy night last night with 11 individual slot command sequences being sent. According to the WW log all were successful.
I had enabled all 7 of my monitors but with a 60 second polling interval so the internal traffic to the inverter would have been considerably reduced but not eliminated.
Looking at the GE log a total of 117 instructions were sent by WW.
4 instructions timed out but were repeated after 2 seconds successfully.
1 inverter offline response was received then received again 2 seconds later but was then successful after a further 9 seconds. (This might have been the inverter being overwhelmed but then the longer retry sequence caught it soon after?)
So it looks like reducing internal traffic and increasing the time between instruction retries has solved the original problem of around 1 in 10 WW command sequences failing - I'll continue to monitor this to pick up on any corner cases or any other difficulties with the reliability of web control of the inverter.