Shake having large data gaps and connection issues

Technical Support
1

Hello,

I am having an intermittent problem with my Raspberry Shake and I would appreciate some help identifying the cause.

The Shake is installed in a temporarily unoccupied house. The house has power and Internet, and the Shake is currently connected to the router over Wi-Fi. I chose this location because it has very low seismic noise, so I would prefer to keep the station there. I also plan to switch it to Ethernet in the future.

The problem is that I am getting data gaps that can sometimes last for several hours.

I have noticed a few different situations:

  • Sometimes Winston Wave stops working, but SeedLink continues to work.
  • Sometimes Winston Wave, SeedLink, rs.local and the helicorder all become unavailable at about the same time, while ZeroTier still shows the Shake as connected.
  • Generally after that ZeroTier also disconnects and the Shake becomes completely unreachable for hours, this coincides with the large data gaps, and when the shake connects back, its when the data starts to appear again on the helicoders.

The gaps shown in Swarm’s helicorder are the same gaps I see through rs.local/heli.

This started happening after the ISP at the house was changed (right now the RS is connected via WiFi to the Router, im planning to change to ethernet). Since then, the Shake has also had problems connecting to the Raspberry Shake servers, although I can still access it remotely through ZeroTier. I am going to contact the ISP to ask about the required connections for the RS server. I was told in another thread that ports 55555, 55556 and 123 may need to be allowed.

What I am more concerned about, however, are the cases where ZeroTier also disconnects. Since the house is currently unoccupied, I cannot easily check the Shake when a gap occurs.

I have attached the log files from the Shake. Could you please take a look and let me know if there is anything indicating:

  • an unexpected reboot or power interruption
  • Wi-Fi or network disconnections
  • Winston Wave or SeedLink stopping/restarting
  • filesystem or storage problems
  • hardware-related errors
  • or anything else that could explain the long gaps

I am particularly interested in whether the logs can distinguish between the cases where only Winston Wave stops, where several services stop while ZeroTier remains connected, and where the entire Shake also disappears from ZeroTier and the data stops being recorded.

I am planning to visit the house during a gap if possible, so I can check the power, router, LEDs, network connection and the physical state of the Shake.

Thank you for your help. Please let me know if you need any additional logs or information.
RSH.RAAC8.2026-08-21T15_26_27.logs.tar (4.2 MB)

Captura de pantalla 2026-08-21 092453
Captura de pantalla 2026-08-21 092453958×957 77.7 KB

2

Hello Jonathan, and thank you for the detailed description of the situation and the logs from the Shake.

Usually, gaps are related to power supply issues, but that doesn’t seem to be the case here. Instead, it appears that the Shake is having out-of-memory errors with associated segmentation faults. These cause the associated data services to fail, then resume when enough memory becomes available again.

The reason behind these errors is not apparent, though. Do you have any other software installed on the Shake? And, if you have kept a tally, do these data interruptions happen more or less at regular intervals (or at regular times), or are they completely randomized?

Thank you.

3

Hello,

Right now, the only additional software installed on the Shake is ZeroTier.

The interruptions don’t seem to happen at regular intervals. They appear to be randomized, although I have noticed that many of them happen during the night. I am not sure if the time of day is actually significant, though.

At the moment, the issue is happening again. I can still access the Shake through SSH and the web panel, but the services are showing as down in the panel:

image
image1920×898 115 KB

Since you mentioned out-of-memory errors and segmentation faults, could these possibly be related to the Shake repeatedly trying to connect to the Raspberry Shake servers? The ISP at this location appears to block or interfere with the Shake’s connection to the Raspberry Shake servers, although ZeroTier and SSH/web access still work.

Would it be useful to temporarily change the Shake to offline mode until I can move it to another location/network, or untill i can ask the ISP to see if they can allow the required connections?

I will also provide the logs from this latest occurrence.

Thank you!
RSH.RAAC8.2026-08-24T15_04_21.logs.tar (3.6 MB)

4

Hello, thank you for the additional logs and the new details.

Since the process that retries a blocked outbound connection is a very light operation, I think it’s unlikely that this is causing the memory issues. However, if we want to fully rule this out, and since you’re planning the visit to the Shake location anyway, I recommend retrieving the instrument and reconnecting it through another network you know is unblocked.

This will also help in future testing/troubleshooting since the Shake is closer to you.

If the memory issues continue, I recommend reburning the Shake OS on a new microSD card to see whether the problem is software-related. If needed, I’ve left the instructions here.

5

Hi again,

I have decided to take it back to the old location, where I can troubleshoot it more easily. Hopefully, next weekend it will be installed at a new location with a stable connection and similar seismic noise levels to the current site.

I’m also planning to construct a seismic vault later this year so I can eventually move the station to a more remote area with little to no human activity.

For now, it is connecting to the Raspberry Shake servers again, so I hope the issue does not come back.

Best,
Jonathan

1 Like