1

I am using a rpi3 with an external usb disk and a 3A power source, but the system seems to turn off after 1-2 days of normal use.

I know that because I cannot access it (ssh or vnc), but there is a red light there.

Then, taking a look at /var/log/messages I think to see when it died, since it stops logging messages 30h ago. Nothing seems out of normal, except for the repetitive (which is also doing it now, just restarted)

Aug 19 05:49:03 raspberrypi kernel: [ 1967.123127] cfg80211: World regulatory domain updated:
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123134] cfg80211:  DFS Master region: unset
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123140] cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp), (dfs_cac_time)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123150] cfg80211:   (2402000 KHz - 2472000 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123158] cfg80211:   (2457000 KHz - 2482000 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123166] cfg80211:   (2474000 KHz - 2494000 KHz @ 20000 KHz), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123175] cfg80211:   (5170000 KHz - 5250000 KHz @ 80000 KHz, 160000 KHz AUTO), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123185] cfg80211:   (5250000 KHz - 5330000 KHz @ 80000 KHz, 160000 KHz AUTO), (N/A, 2000 mBm), (0 s)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123193] cfg80211:   (5490000 KHz - 5730000 KHz @ 160000 KHz), (N/A, 2000 mBm), (0 s)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123201] cfg80211:   (5735000 KHz - 5835000 KHz @ 80000 KHz), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.123209] cfg80211:   (57240000 KHz - 63720000 KHz @ 2160000 KHz), (N/A, 0 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151591] cfg80211: Regulatory domain changed to country: ES
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151609] cfg80211:  DFS Master region: ETSI
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151620] cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp), (dfs_cac_time)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151630] cfg80211:   (2400000 KHz - 2483500 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151640] cfg80211:   (5150000 KHz - 5250000 KHz @ 80000 KHz, 200000 KHz AUTO), (N/A, 2301 mBm), (N/A)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151649] cfg80211:   (5250000 KHz - 5350000 KHz @ 80000 KHz, 200000 KHz AUTO), (N/A, 2000 mBm), (0 s)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151658] cfg80211:   (5470000 KHz - 5725000 KHz @ 160000 KHz), (N/A, 2698 mBm), (0 s)
Aug 19 05:49:03 raspberrypi kernel: [ 1967.151665] cfg80211:   (57000000 KHz - 66000000 KHz @ 2160000 KHz), (N/A, 4000 mBm), (N/A)

Maybe this is completely unrelated. In any case, what can I do to see what's causing this instability of the system? Is it likely to a not good enough power supply?

Improve this question
8
  • Are you connecting to the Pi over WiFi or Ethernet? It could be that only WiFi is dying. I suggest connecting the Pi to a monitor and keyboard to determine if the machine is still running or not.
    – Steve Robillard
    Aug 19, 2016 at 5:58
  • Logs in /var/log/messages stopped at a certain point, and didn't log anything else until I manually unplug and plug again the rpi (30h later). To me this indicates that the whole machine stopped doing anything (doesn't it?)
    – GWorking
    Aug 19, 2016 at 6:07
  • 1
    It might but by connecting the monitor you can be sure. The most common cause for this would be power issues. Don't go by what is printed on the label you need to actually measure the voltage onthe board. When ddebugging an implication can lead to hours of chasing the wrong cause.
    – Steve Robillard
    Aug 19, 2016 at 6:09
  • I would say its always better to avoid loading raspberry pi with external storage devices. I have a belkin router (N 600). I have tied my external storage to this router. This makes the external storage device accessible on my system.
    – Varad A G
    Aug 19, 2016 at 6:34
  • @VaradAG there are many guides claiming that with a 3A it is perfectly doable to have a rpi (3 in this case) with a standard external USB hard disk
    – GWorking
    Aug 19, 2016 at 12:06

1 Answer 1

Reset to default
0

I had the same issue, but it was happening once in a few hours. In my case, it was the power supply fault (labeled as 2.1A but I'm afraid it didn't get any close to that).

  1. Check your power adapter, try another one. I use bulky ones that are dedicated to tablets.
  2. Your USB disk might draw too much energy, so maybe try using powered hub
  3. If not, it might be caused by the voltage drop in the electrical grid inside your building (maybe switching on electric cooker or kettle, AC, some kind of engine?) - the drop might be long and deep enough to cause Pi to freeze. If you have UPS, try connecting your power supply to it.
  4. You might also check if your system is okay. Try another SD card with freshly installed OS.
  5. This could also be caused by some damage to your Pi or the external disk that causes some kind of short circut (cold solder joint / dry solder joint). Change them and check if the problem is gone.
Improve this answer
5
  • Yes, I've bought another supply hopefully more true 3A. That said, with 3A a standard usb external disk should go ok, or at least there's plenty of users claiming this out there. I also know it's not a WIFI failure since I have a script that resets the rpi and sends me a message in that case, so I know the system is not alive. Other than that, my question was more indicated towards a software approach to the problem, perhaps there are some logs that could point me to where the problem can be, perhaps I can monitor every 5 minutes the amperage consumed, if that can be monitored by software
    – GWorking
    Aug 19, 2016 at 12:03
  • Anything that prevents me from applying your measures in a blind manner. Can it be a temperature problem? How can I detect / monitor it? Is there anything the rpi can tell me about what's happening to it? Or all I can do is what you propose, this is, apply all potential solutions sorted by possibility but not based on any evidence
    – GWorking
    Aug 19, 2016 at 12:05
  • You can't really monitor system "A" without separate system "B". Software solution? Reflash OS. Same goes with regular PC's. When CPU stops, nothing can help you.
    – Mark
    Aug 19, 2016 at 13:09
  • Well, that was the doubt that I had, so thank you :)
    – GWorking
    Aug 19, 2016 at 19:07
  • I have changed the supply for an amazon well-reviewed 3A (the previous one was also of 3A), and it is actually crashing earlier. Indeed, for more than a week the previous source has gone ok, I have the impression than when the hard disk has to work then there's too much current in game. However, since 3A should be plenty for a rpi and a hard disk, I need to discard other things. Based on raspberrypi.org/forums/viewtopic.php?f=91&t=34994 I am now monitoring the temperature every hour, let's see what do I get
    – GWorking
    Aug 29, 2016 at 20:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.