StorageError: The device storage is full while learning and workaround

[chazeon]

I recently got a new Broadlink RM mini 3 from Amazon US, the device type is 0x5F36. With python-broadlink, I am able to "learn" a few IR code, it is working fine. After a few codes learned, the device starts to complain "The device storage is full" error. This problem cannot be resolved by resetting the device, and I am pretty desperate at this point.

On deciding whether I will need to return the device, I realize that: after using enter_learning() function, the device has white indicator on, which means it indeed reads the IR code, but the error only comes when I try to use check_data() function. Further tests indicate that as long as I commented out the check_error() statement, the check_data function can actually correctly reporting IR code.

So I assume the "learning" functionality is actually writing IR codes into the device's memory, which, from what I see, is very much limited. Since there is no "erase" option given, I assume a lot of users are going to encounter the same problem. And, since most users of python-broadlink only need to read out the IR code instead of really using the learning function, I think it would be great if an option for temporarily disabling the error check for full device storage could be given.

[fiercedruid]

I'm having the same issue here. I tried commenting out the line you reference and the script still crashes.

[chazeon]

I'm having the same issue here. I tried commenting out the line you reference and the script still crashes.

Which script are you using? broadlink_cli? I have issues with broadlink_cli as well.
But I can work with the most basic commands. Try the following lines in the Python interactive shell:

First find your device and enter learning mode

import broadlink

device = broadlink.discover()
devices.auth()
device.enter_learning()

Now the LED indicator should be white, press the button now. After that, run:

packet = devices.check_data()
print(packet.hex())

This should show the IR packet in hex format.

[digitlength]

But I can work with the most basic commands. Try the following lines in the Python interactive shell:

Can you confirm whether to try these commands in the Python console after commenting out line 74 in exceptions.py, or doesn't it require commenting out that line at all?

[chazeon]

But I can work with the most basic commands. Try the following lines in the Python interactive shell:

Can you confirm whether to try these commands in the Python console after commenting out line 74 in exceptions.py, or doesn't it require commenting out that line at all?

Here I mean after commenting out line 74.

But actually, in my test, the first portion works even without commenting out the line. The check_data method in the second part is giving out StorageError.

[digitlength]

But actually, in my test, the first portion works even without commenting out the line. The check_data method in the second part is giving out StorageError.

So just to check: are you commenting out only line 74, or also the whole check_error() function?

[mjeshurun]

Hello friends,
Good to know I'm not the only one suffering from the "The device storage is full" issue.
I have an RM4 Pro and when I try to learn RF codes using python-broadlink or broadlink-mqtt I'm seeing the "device storage is full" error for both learning methods.

By the way, I don't know if it's a problem, but my RM4 Pro is showing a red LED when its in learning mode (not white LED).

Here is another thread which I tried using to troubleshoot the issue (with the kind help of @hafniumzinc), but, so far, with no success:
https://community.openhab.org/t/broadlink-bestcon-rm4c-mini-via-mqtt/98268

Have you found any new leads to solve the problem?

[mwargan]

+1 on facing this issue. Interestingly I noticed that when you learn the command from Home Assistant - works fine. When using the broadlink_cli here, it returns the full storage error.

[felipediel]

This is not a bug, this is the expected behavior. The applications that import this library should handle this exception using a try except clause. Check the Broadlink integration in Home Assistant for more implementation details.

[chazeon]

This is not a bug, this is the expected behavior. The applications that import this library should handle this exception using a try except clause. Check the Broadlink integration in Home Assistant for more implementation details.

Yes, I understand it is reasonable in the API. But the boradlink_cli has not implemented the try-except clause. Which makes the learning functionality the CLI totally unusable for me. The boradlink_cli is supposed to be working for end-users right?

[mcgheffe]

Is the the error "The device storage is full" supposed to be fixed now? I have the pulled the latest which includes the fix to broadlink_cli to handle the StorageError exception, but I still encounter the error when I try to do anything with my RM3 mini.

[felipediel]

Please make sure your code is correct. This is our implementation in Home Assistant. You can use it as a template.

[mcgheffe]

I see the difference now. In the current broadlink_cli python, the call to dev.check_data() is wrapped in an 'try', but I'm getting the exception from dev.enter_learning(), which is not wrapped in exception handling:

Traceback (most recent call last):
  File "./broadlink_cli", line 132, in <module>
    dev.enter_learning()
  File "/usr/local/lib/python3.6/dist-packages/broadlink-0.15.0-py3.6.egg/broadlink/remote.py", line 38, in enter_learning
    check_error(response[0x22:0x24])
  File "/usr/local/lib/python3.6/dist-packages/broadlink-0.15.0-py3.6.egg/broadlink/exceptions.py", line 137, in check_error
    raise exception(error_code)
broadlink.exceptions.StorageError: [Errno -5] The device storage is full

However, wrapping dev.enter_learning() in a 'try' block doesn't help. It must fail to enter learning mode since it just times out later reading data:

### Ignoring Storage Error ###
Learning...
No data received...

I also can't send an IR code I've already learned, so there's a fundamental problem executing any useful operation due to the storage error.

[felipediel]

You are probably using the wrong class. Try rm4. You should also consider using discover() or gendevice() to generate the device correctly.

[mcgheffe]

Thanks for the pointer. You are right. Changing from using the "rm" class to using the "rm4" class works! Glad to finally get this working!

[mcgheffe]

One more note just to complete my view of things. My "old" RM3 mini still requires the old "rm" class (old firmware being "v57"). The new RM3 minis I have purchased recently do not work with the "rm" class, they require the use of the "rm4" class (new firmware being "v44057"). Thanks for the help sorting this out.

[felipediel]

Yes. RM mini 3 0x5f36 requires special payload headers. It communicates like RM4 series devices.

[mjeshurun]

Thanks for the pointer. You are right. Changing from using the "rm" class to using the "rm4" class works! Glad to finally get this working!

Hi friends,
Based on the recent posts in this thread am I correct to assume that the "device storage is full" issue has been fixed?
I own an RM4 Pro and tried to learn RF codes, but I'm still getting the "storage full" error message.
I could use your help in solving this issue, since I'm quite a noob in these coding matters 🙏

[KTibow]

@mjeshurun what device class are you using? Try a different one, like rm4 instead of rm?

[mjeshurun]

@mjeshurun what device class are you using? Try a different one, like rm4 instead of rm?

Thank you for your reply @KTibow ,
Where do I change the device class to rm4?

When I ran the following command python-broadlink discovered my device as an rm4:

pi@raspberrypi:~/python-broadlink/cli $ ./broadlink_discovery --timeout 10 --dst-ip 192.168.1.6
Discovering...
###########################################
RM4
# broadlink_cli --type 0x61a2 --host 192.168.1.6 --mac 24dfa7b9c689
Device file data (to be used with --device @filename in broadlink_cli) : 
0x61a2 192.168.1.6 24dfa7b9c689
temperature = 0.0

[felipediel]

What command are you using to learn the code?

[mjeshurun]

What command are you using to learn the code?

I'm following this set of commands:

python3

>>> import broadlink

>>> devices = broadlink.discover(timeout=5)

>>> print(devices)

>>> devices[0].auth()

>>> devices[0].enter_learning()

After the last "enter_learning" command I get the error message, so I cant continue to the next commands that are supposed to be:

>>> packet = devices[0].check_data()

>>> print(packet.hex())

>>> devices[0].send_data(packet)

[felipediel]

When you call devices[0].check_data(), you need to catch the exceptions with a try: except clause. Check the CLI implementation for more details.

[mjeshurun]

When you call devices[0].check_data(), you need to catch the exceptions with a try: except clause. Check the CLI implementation for more details.

Thank you for the suggestion, @felipediel .
Can you please explain what you mean by catching the exceptions with a try: except clause?
I don't really know how to code, so I don't know what to do. I can follow detailed guides, but I don't know how to understand coding functions that might be naturally understood by a developer.
If you are able, I would really appreciate a clarification of what steps I need to take to catch the exceptions 🙏🙏

[felipediel]

The try: except clause is similar to an if statement, but we ask for forgiveness instead of asking for permission in order to improve performance.

If statement (asking for permission)

if code_received:
    device.check_data()
else:
    print("No code received")

Try except (asking for forgiveness)

try:
    device.check_data()
except ReadError:
    print("No code received")

This code snippet is exactly what you are looking for.

[mjeshurun]

The try: except clause is similar to an if statement, but we ask for forgiveness instead of asking for permission in order to improve performance.

If statement (asking for permission)

if code_received:
    device.check_data()
else:
    print("No code received")

Try except (asking for forgiveness)

try:
    device.check_data()
except ReadError:
    print("No code received")

This code snippet is exactly what you are looking for.

Thank you 🙏
To double check I understand what you mean.
My next step is to open the broadlink_cli file with a text editor and check if the commands you wrote (and those highlighted in the code snippet) are in the broadlink_cli file verbatim?

Or do I need to add the commands you wrote somewhere into the broadlink_cli file?

[felipediel]

If you just want to learn a code, you can use the CLI:

broadlink_cli --device "0x2712 192.168.0.16 aaaaaaaaaa" --learn

If you want to create a Python script, you can refer to the CLI to understand how the library works and how its functions should be implemented.

[mjeshurun]

@felipediel
Do you mean in python3 I need to to use this learn command:
broadlink_cli --device "0x61a2 192.168.0.6 24:df:a7:b9:c6:89" --learn

And not use the earlier learn command I mentioned before? ( devices[0].enter_learning() )

Because doing this didn't work. I got a SyntaxError: invalid syntax message.

[KTibow]

@mjeshurun they mean to run it from the command line, possibly with python3 -m before it.

[mjeshurun]

@mjeshurun they mean to run it from the command line, possibly with python3 -m before it.

Thanks @KTibow .
I tried running this command python3 -m broadlink_cli --device "0x61a2 192.168.0.6 24:df:a7:b9:c6:89" --learn
in terminal at the python-broadlink folder and also in the cli folder, but it didn't work.
Then I tried running a python3 shell and run the learn command in the shell in two ways:
python3 -m broadlink_cli --device "0x61a2 192.168.0.6 24:df:a7:b9:c6:89" --learn
and
-m broadlink_cli --device "0x61a2 192.168.0.6 24:df:a7:b9:c6:89" --learn
But both commands gave me the SyntaxError: invalid syntaxmessage.
I guess I'm still doing something wrong due to not understanding something.

[KTibow]

As for the first one, what do you mean? Did it give any errors? You need to replace the 0x61a2 192.168.0.6 24:df:a7:b9:c6:89 with your 0xdevicesig deviceip devicemac

[mjeshurun]

As for the first one, what do you mean? Did it give any errors? You need to replace the 0x61a2 192.168.0.6 24:df:a7:b9:c6:89 with your 0xdevicesig deviceip devicemac

This is the command I entered in terminal and the error message I got.

pi@raspberrypi:~/python-broadlink/cli $ python3 -m broadlink_cli --device "0x61a2 192.168.0.6 24:df:a7:b9:c6:89" --learn
/usr/bin/python: No module named broadlink_cli

Thanks for double checking. Yes, my device info (type, ip, mac) are mentioned in the command 0x61a2 192.168.0.6 24:df:a7:b9:c6:89.

[felipediel]

./broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn

or

python3 broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn

Do not use : in the MAC address.

[mjeshurun]

./broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn

or

python3 broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn

Do not use : in the MAC address.

Thank you. I tried both commands, but I got the same error message.

pi@raspberrypi:~/python-broadlink/cli $ python3 broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn
Traceback (most recent call last):
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 369, in send_packet
    response = cs.recvfrom(2048)
socket.timeout: timed out

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "broadlink_cli", line 105, in <module>
    dev.auth()
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 270, in auth
    response = self.send_packet(0x65, payload)
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 374, in send_packet
    raise exception(0xfffd)
broadlink.exceptions.DeviceOfflineError: The device is offline

pi@raspberrypi:~/python-broadlink/cli $ ./broadlink_cli --device "0x61a2 192.168.0.6 24dfa7b9c689" --learn
Traceback (most recent call last):
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 369, in send_packet
    response = cs.recvfrom(2048)
socket.timeout: timed out

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "./broadlink_cli", line 105, in <module>
    dev.auth()
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 270, in auth
    response = self.send_packet(0x65, payload)
  File "/usr/local/lib/python3.7/dist-packages/broadlink/__init__.py", line 374, in send_packet
    raise exception(0xfffd)
broadlink.exceptions.DeviceOfflineError: The device is offline

[felipediel]

I think the IP address is wrong.

[mjeshurun]

I think the IP address is wrong.

Wonderful! You were right!!!
I didn't notice I made a small mistake in the IP address. Sorry for this foolish mistake.

My next step was to record my appliances wireless remotes' codes. However, the RM4 Pro didn't recognize any RF codes, and, as such, it returned a "No data received..." message.

pi@raspberrypi:~/python-broadlink/cli $ python3 broadlink_cli --device "0x61a2 192.168.1.6 24dfa7b9c689" --learn
Learning...
No data received...

On the other hand, when I tried to record IR codes the RM4 Pro properly recorded them:

pi@raspberrypi:~/python-broadlink/cli $ python3 broadlink_cli --device "0x61a2 192.168.1.6 24dfa7b9c689" --learn
Learning...
2600680112000249610001270e130e330f1111101110110f1110110f111011301110110f1130111110100f320f330f32113011301110110f1110110f11101110110f1110111010110e120f120e1210111110110f1110110f1110110f120f1110110f1110110f1111101011100e130e120e130f110f32113111301130116162000125113011101110111010110e120e130f110f1211301110110f1130111011301130113110310f320f330e120f12110f11100f12110f1110110f11100f1111100f12110f0f130e120f120e120f120f11101111100f1110110f1110110f120f110f120f1110110f120f120e130e120e130e120f636100012710310f1110110f1110110f120f110f120f1110320f120e120f330e330f1110320f110f32103110320f320f320f330e330e330f120f1110110f320f320f3210110f120f110f120f1110320f320f330e120f3210110f1110320f320f1110110f120f110f120f120e130e330e330e330f3210000d050000000000000000000000000000
Base64: b'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'

Am I missing a necessary step before recording RF codes?

[felipediel]

You can use python3 broadlink_cli --device "0x61a2 192.168.1.6 24dfa7b9c689" --rfscanlearn.

[mjeshurun]

You can use python3 broadlink_cli --device "0x61a2 192.168.1.6 24dfa7b9c689" --rfscanlearn.

YAYAYAYAYAY!!! This command did the trick!!! :)))))
Now I need to learn the RF codes and configure the OpenHAB binding :)))
I'm so glad I'm able to continue to the next step.
Thank you so much for your help @felipediel @KTibow 🙏🙏🙏

[felipediel]

Closing, since it is not a problem to be solved in our code base.

[rodrigogml]

Sorry I reopened this thread. The problem still happens, but I think I can help to replicate. And I believe it is a code problem.

Hope this help to solve. I spent my afternoon trying to make it work and getting this error. I notice when we started to learn any code, and don't press any button on remote, the error popups in screen (about 5 seconds after the command) while the RM4 Mini still with the white led on.

But, if you press the remote control button just after entered the learning mode, it works beautifully.

So to replicate the problem, just enter the learn mode and just wait 5 seconds.

I'm a developer but zero skills with python. So, I suppose, the problemas has something about the delay between entering the learning mode and "grab" the code from the device while it still not ready yet.

And the error about full storage is just a "bad catch" somewhere, or the "bad code" returned crashes the write method.

Hope it drives for the solution, and btw REALLY THANKS for all the effort spent.

[felipediel]

Hi. We are catching the StorageError here. The problem does not exist in our code base. It is an implementation problem in client applications, which should use the CLI as a starting point to implement our functions correctly.

Edit: I just realized we need to improve our documentation to inform about this behavior.

[rodrigogml]

You're right! My mistake, I'm sorry!

I was able to detect the problem using the library directly on python console, but the treatment is indeed in the other project.

*By problem I meant ask to check_data() before pressing the button.