Skipp... Our application is on DOCSIS network... Closed network... Works well for the roughly 125k devices on it...
Auto polling for mapping is about every 90 seconds - refreshing the map manually causes a manual poll which takes about 10 seconds to complete. Worst case we're looking at about a 3 minute delay between poll time and map display if it misses the refresh. Once a minute per modem on random intervals... Not sure about your particular issue with PG&E - the majority of the timeframe issues I've had with regulated utilities like power and phone have been as a result of PUC mandates... Anything sooner than 3 days would be an unfair advantage over competition LOL. Sorry for the segway out of Repeaterland... On Tue, Jan 12, 2010 at 10:28 AM, skipp025 <[email protected]> wrote: > > > > > Reverse the logic - if it's a non-responder (no telemetry > > after n minutes), it shows up on the map as a customer out. > > OK, but it can't be practical in this application (by the sheer > number of units in service) to poll individual specific locations > within n-minutes... and and that amount of time is going to > probably be hours at best. > > > We currently have a Google maps-based outage tracking > > system that places a green dot for working modem, yellow > > for modem syncing and red for offline... > > and the update time for a specific customer location is > estimated to be _____ ? > > > > Since we have SNMP polling at around once a minute, along > > with bandwidth monitoring in place, often times we can find > > out about an outage before the customer can even find their > > phone to call us. > > Once a minute at a specific location? > > > It's just another tool we use of many... > > I would consider it very useful... but it doesn't shorten the > Bureaucracy. It still has taken 3 days for PG&E to turn on > the service I requested (not counting the 4 non work Sunday). > I hope to have the service active when I visit the site later > today... > > cheers, > s. > > >
