Repository navigation
Reterminal E1003 freezes after some hours #116
Description
Activity
Hi thanks for using OpenDisplay. Overall stability should be great. We have noticed some issues with BLE communication particularly when using a Bluetooth Dongle and ESP32 displays. These have been addressed in the latest firmware. Please flash to 2.23.
You can also access the on-device debug log via:
Click on Connect, select the device from the COM ports (while connected to USB). Then click on LOG and it will show the live log from the device.
Thank you.
Thanks. I updated the firmware and give it a try. Logs are difficult because the device is only connected via USB when it's getting a firmware update.
What I can see currently after a few hours:
Sometimes I get the following error (trace from the HA automation):Error: Errors occurred for one or more devices: 8aeb60c7e4550a3e2568bbf6dedfe1a4: Failed to upload to the display: PIPE_WRITE aborted: MAX_RETX (48) exceeded
sometimes
Error: Errors occurred for one or more devices: 8aeb60c7e4550a3e2568bbf6dedfe1a4: Failed to upload to the display: PIPE_WRITE aborted: no ACK progress after 3 PTO probes
but 1 try later it works again.
Accessability is good (last seen changes every minute), RSSI around -55
Thanks this is helpful. Can you reflash to latest firmware 2.23? It should fix this. Let me know if it works.
Thanks this is helpful. Can you reflash to latest firmware 2.23? It should fix this. Let me know if it works.
That is already under 2.23.
I see the same issue.
Is deep sleep enabled?
Is deep sleep enabled?
No.
I use deep sleep. I can try to debug it, but we timestamps in the log would be really useful to make that easier.
Just to underline the behaviour: Sometimes the automation is executed successfully, sometimes it shows errors (this time "not connected")
But last seen updates regularily (which is an improvement with 2.23)
I added better logging support which has a timestamp since boot. If you have firmware with logging which includes timestamps you can boot the device and note the time and if you display a clock on the display you will know at which time it stopped working. By that you can figure out where in the log to look at.
I will do that next week.
@riesi77 do you have encryption/password protection on? I have found some potential failure modes on the Firmware side.
No. But (running 2.23) the device did not freeze any more. Just every third or fourth automation fails. Will try updating to the newest FW in the next days.
I installed 2.25.1 yesterday and added a helper to better visualize if the automation fails or not. 1 is successful, 0 is failed:
The errors:
Just to followup there are some known BLE proxy issues introduced in ESPHome 7.x and later. These may be related as we are seeing multiple similar reports. Still waiting for a resolution. There is a merged pill request.
i use my E1003 with deep sleep and not have this issues. just for info
I am closing this since it might has been cased by another issue. If it persists, please open it again
I have
a Reterminal E1003
with Firmware 2.20
with Home Assistant (On Docker, using ESP32 Bluetooth Proxy) using Integration 3.0.0-beta.7
an Automation running every 16 Mins that renders a quite complex dashboard using OpenDisplay.DrawCustomImage
After several hours (sometimes 8, sometimes 10) the Reterminal gets unavailable, does not refresh. I need to switch it off and on again and then it works again for some hours.
Bluetooth RSSI is around 75 dBm
How can I trace down the error? Is there some logging possible on firmware level (I guess it's not on the integration level)? I could restart the device every few hours, but the integration does not provide a reboot action. Would be great to have that.
I see a lot of technical pull requests, but it's not clear for me what the benefit of those is. I guess stabilization in general. Is the project already mature enough to allow the devices run long term without issues or are issues like mine still expected because the project is still too young?