[Bug]: T1000e losing programing

[madeofstown]

Category

Other

Hardware

Seeed Card Tracker T1000-E

Firmware Version

2.5.16 and 2.5.15

Description

I really like the T1000e for it's features and battery life but there is something strange going on. Within the last 2 weeks (with current firmware 2.5.16 as well as 2.5.15) I've had the T1000e lose its programming over night while charging 2 times. I've also noticed that it will restart itself a few times during the day playing the telltale startup tune on its buzzer. Because of the time when this is happening (overnight while I'm sleeping) I am not able to provide any logs.

Relevant log output

No response

[thebentern]

Please update us if you are able to get some logs.
A couple of questions with setup, do you use MQTT w/ client proxy or are you on a busy mesh (i.e. >100 nodes)?

[madeofstown]

Quick update. First, not using MQTT at all but my node database usually sits at 80 nodes (I believe this is maxed out for an nRF chip)

The first time I lost programming about a week ago it seemed to be just the node database (name was reset but most settings stayed) that was wiped. When it happened again last night everything is now gone.

[garthvh]

Are you connected via BLE to an android phone?

[madeofstown]

Are you connected via BLE to an android phone?

Yes, this is my on-person node so I try to make sure it's always connected to my phone.

[todd-herbert]

I believe I'd heard @erayd mention that his T1000E had once run out of filesystem space, even after the Node DB size was dropped to 80.

[madeofstown]

I believe I'd heard @erayd mention that his T1000E had once run out of filesystem space, even after the Node DB size was dropped to 80.

I've also experienced losing programing several times on a RAK4631 solar node that I've been testing. I wonder if 80 nodes in the DB is still too many for nRF based nodes

[garthvh]

I believe I'd heard @erayd mention that his T1000E had once run out of filesystem space, even after the Node DB size was dropped to 80.

I've also experienced losing programing several times on a RAK4631 solar node that I've been testing. I wonder if 80 nodes in the DB is still too many for nRF based nodes

There is an Android BLE disconnect when out of range bug that seems to be the source of the lost user proto that we have not been able to track down. I have seen the memory wipe while with a device using iOS but only when on the root topic of mqtt for 24 hours +

[madeofstown]

I believe I'd heard @erayd mention that his T1000E had once run out of filesystem space, even after the Node DB size was dropped to 80.

I've also experienced losing programing several times on a RAK4631 solar node that I've been testing. I wonder if 80 nodes in the DB is still too many for nRF based nodes

There is an Android BLE disconnect when out of range bug that seems to be the source of the lost user proto that we have not been able to track down. I have seen the memory wipe while with a device using iOS but only when on the root topic of mqtt for 24 hours +

Interesting, I don't think that's my problem since I keep the T1000e clipped to my shirt and my phone in my front pocket and at night they charge on the same nightstand. I do notice that I occasionally need to reconnect the node during the day but I didn't think it was a range issue.

[b8b8]

Maybe not related specifically but i just had the T1000E GPS lockup for about 13 hours over night and refuse to get GPS lock after running fine for days. ive since rebooted it and am trying to get it to lock again.

[b8b8]

Im having a RAK have a hard time maintaining its name as well as it just randomly wiped its memory today, including losing its custom bluetooth PIN. It is a "router" node but has been set to client_mute for the last few weeks since its off the mountain right now. I upgraded to 2.5.17 this morning and started having the issues. This seems to be similar to the T1000E problem. This node was NOT CONNECTED to any phones for the last month (nor this morning). Node database was only 35 nodes at the time.

[Firecul]

I haven't experienced that myself and I am going out of range of my T1000e (2.5.15) frequently. I do however keep my node list quite empty. If I haven't seen a node in a few days I'll normally clear it off. It's possible that has protected me from this bug.

[b8b8]

I haven't experienced that myself and I am going out of range of my T1000e (2.5.15) frequently. I do however keep my node list quite empty. If I haven't seen a node in a few days I'll normally clear it off. It's possible that has protected me from this bug.

What phone are you using? It seems to be more prevalent on Android, specifically Pixel phones.

[Firecul]

What phone are you using? It seems to be more prevalent on Android, specifically Pixel phones.

I've been switching between my Pixel 6 Pro and iPhone SE 2020 to use with that node for a few months. I've been running it more with the iPhone as I can more easily grab the metric logs from there while I'm experimenting.