CD108 Geschrieben July 4, 2014 at 10:11 Geschrieben July 4, 2014 at 10:11 I have 3 master bricks with wifi extensions measuring temperatures etc around my house. Lately I've noticed that the bricks are freezing and becoming unreachable. The only thing I can do to connect is reset the master bricks. The bricks are connected to dedicated wifi and are set to use static ips. Everything is using the latest firmware. I am using c# and connecting using the 'rugged' approach i.e. using enumerate callbacks. Failures are not all 3 at once - one brick freezes then the others at random times. Sometimes they run for days, but lately they have only been working for hours. Any ideas? Zitieren
remotecontrol Geschrieben July 5, 2014 at 10:12 Geschrieben July 5, 2014 at 10:12 You have one C# application which connects to three bricks simultaneously or do you disconnect the others before you connect to one brick (only one is connected at a time)? Can you still "ping" the bricks if they are unreachable via API? Are the bricks located nearby other electric devices? Can you see anything specific in the access point's log? Do you have the possibility to connect a brick to USB if the brick does not respond to wifi and check if brickv finds the usb connected device? Zitieren
CD108 Geschrieben July 5, 2014 at 11:47 Autor Geschrieben July 5, 2014 at 11:47 One application which connects to all 3 using different ipconnection objects (all running at the same time). Access point shows the devices as not connected when they freeze. Devices are no longer pingable. Devices are not near any other electronics. I have not tried connecting via usb as they are powered through the usb ports. Disconnecting would result in a power cycle. It looks like either the brick or wifi are hanging and never recovering Zitieren
remotecontrol Geschrieben July 5, 2014 at 13:34 Geschrieben July 5, 2014 at 13:34 Are you sure that your application doesn't deadlock? If you receive data from three bricks via callbacks you must properly synchronize your application threads. It might be that in case of an application deadlock the stack cannot send data any more and then has an internal overflow (just a guess). How many callbacks do you typically get per second? Zitieren
CD108 Geschrieben July 5, 2014 at 16:42 Autor Geschrieben July 5, 2014 at 16:42 Callbacks are set to between 10 and 60 second intervals over 8 bricklets so less than 1/sec. High speed measurement isn't required for the application. I call all 3 connections on a single thread with each different type of bricklet directed to common callbacks, however it doesn't appear to be the problem - sometimes 1 of the bricks freezes and the others continue to operate. I'll try separating them out to see if i still get the problem Zitieren
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.