Automatically stops recording due to low disk space in GUI v1.0.2

[chiyu1203]

Dear OpenEphys team,
I understand a function of automatic stop recording was introduced in the GUI v1.0.2. After I upgraded the GUI, I found out that OpenEphys GUI would stop the recording when there is still around 50 GB in the local space (for 2 consecutive weeks).
50GB does not sound a lot but is still enough in our setup (using RHD2164, recording at 30k/s, one hour recording takes at most 20GB), and especially I always check notices from the Recording Node to estimate how much space is still free, so I would prefer to disable this "automatic stopping recording" function and manage local space by myself. Up to date, my approach to disabling that function is simply downgrade the GUI back to v1.0.1. Is there any other thing you would suggest me to do (I have not tried moving previous recording sessions to online cloud the same time when I am recording the next session. Dont know if that would affect anything or not.....)

[anjaldoshi]

Hi @chiyu1203,

Thanks for reporting this! To help diagnose what's happening, I need a bit more clarification:

When does the recording stop?

• Does it stop immediately when you start recording?
• Or does it stop while recording is actively running?

The behavior depends on when the stop occurs:

1. If recording stops while actively running: This feature was actually introduced back in v0.6.3, not v1.0.2. It automatically stops recording when less than 5 minutes of recording space remains (calculated based on your data write speed and available disk space). You should see the alert: "Recording stopped. Less than 5 minutes of disk space remaining."

2. If recording stops immediately upon starting: This is what's new in v1.0.2. Each Record Node now checks if files can be opened successfully at the start of recording and stops immediately if there are disk space or permission issues. You would see: "Unable to start recording. Please check the console for errors." in the Message Center.

Could you share your console log? This will help us confirm exactly why the recording stopped and whether this is the expected behavior or a potential issue with the disk space threshold calculation.

Regarding the 50GB threshold: The disk space checking logic calculates the remaining recording time dynamically based on the current data rate and available space (see the implementation [here]). If your data rate fluctuates significantly during recording, the checker might be calculating a more conservative estimate than expected, potentially triggering the stop earlier than necessary. With 50GB free and ~20GB/hour usage, you should theoretically have 2.5+ hours of space, so if it's stopping with that much available, the data rate calculation might be the culprit.

Please share your console log so we can verify what's happening!

[chiyu1203]

Hi @chiyu1203,

Thanks for reporting this! To help diagnose what's happening, I need a bit more clarification:

When does the recording stop?

• Does it stop immediately when you start recording?
• Or does it stop while recording is actively running?

The behavior depends on when the stop occurs:

1. If recording stops while actively running: This feature was actually introduced back in v0.6.3, not v1.0.2. It automatically stops recording when less than 5 minutes of recording space remains (calculated based on your data write speed and available disk space). You should see the alert: "Recording stopped. Less than 5 minutes of disk space remaining."
2. If recording stops immediately upon starting: This is what's new in v1.0.2. Each Record Node now checks if files can be opened successfully at the start of recording and stops immediately if there are disk space or permission issues. You would see: "Unable to start recording. Please check the console for errors." in the Message Center.

Could you share your console log? This will help us confirm exactly why the recording stopped and whether this is the expected behavior or a potential issue with the disk space threshold calculation.

Regarding the 50GB threshold: The disk space checking logic calculates the remaining recording time dynamically based on the current data rate and available space (see the implementation [here]). If your data rate fluctuates significantly during recording, the checker might be calculating a more conservative estimate than expected, potentially triggering the stop earlier than necessary. With 50GB free and ~20GB/hour usage, you should theoretically have 2.5+ hours of space, so if it's stopping with that much available, the data rate calculation might be the culprit.

Please share your console log so we can verify what's happening!

Dear @anjaldoshi
Unfortunately, I did not save any console log and screenshot because the incidents were too painful. The only useful information is that at the first time, the recording stopped at around 24 mins; at the second time, it stopped at around 4 mins. I did not remember any messages about checking the console for errors so initially I thought the local disk space was the problem, and started removing unnecessary software in the computer.
Since I downgraded back to v1.0.1, I have not had this issue so far.
Note: the recording computer is Windows 11 25H2, CPU: Ryzen 5 5600x, Disk: Samsung SSD 980 500GB, GPU: GeForce GTX 1660 Super. The Open Ephys Acquisition Board was bought in 2023

[anjaldoshi]

Hi @chiyu1203, could you please try v1.0.2 using the portable .zip binaries (extract to a separate folder so it will not overwrite your v1.0.1 install) and do the following:

• Run the portable Open Ephys executable (no install).
• Use the same Record Node settings you normally use (sample rate, channels, save path).
• Start a dummy recording (~30 minutes).

If the recording stops, please tell us:

• Whether it stopped immediately or after some time, and whether it is reproducible (happens every run).
• The exact alert message that appears.

If you can also paste the console output, that would be very helpful. This will help us determine whether the issue is specific to GUI v1.0.2 or was an anomaly that caused recordings to stop with v1.0.2 previously. Thanks.

[chiyu1203]

Hi @anjaldoshi, I have run 10 tests in total (with some adjustments, for example, recording in a saline bath; did not have camera sending TTL pulse at 100 Hz). To my surprise, I can not reproduce the low-disk issue (I started with either local disk space 30 GB or 50 GB, and recorded for 1 hour). I wonder whether fluctuating voltage signals from the brain, or some computer-related issues makes the estimation of low local space inaccurate.

The only time for me to get a warning and the GUI stopped my recording is when I continued to record up to 2 hours. The local space had only 5 GB left, which I think it's a reasonable warning.
Below is the message from the Console and the screenshot.

[open-ephys] Open Ephys GUI v1.0.2 (Plugin API v10)
[open-ephys] JUCE v8.0.7
[open-ephys] Operating System: Windows 11
[open-ephys] CPU: AMD Ryzen 5 5600X 6-Core Processor (12 core)

[open-ephys] Audio device name: B27T-7 Pro (NVIDIA High Definition Audio)
[open-ephys] Audio device type: Windows Audio
[open-ephys] Audio output channels: 2
[open-ephys] Audio device sample rate: 44100
[open-ephys] Audio device buffer size: 1024

[open-ephys] Signal chain cleared.
[open-ephys] Creating processor with name: Acquisition Board
[acquisition-board] Searching for Opal Kelly Acquisition Board...
[acquisition-board] No Opal Kelly Acquisition Board found.
[acquisition-board] Searching for ONI Acquisition Board...
[acquisition-board] Board opened successfully.
[acquisition-board] ONI Library version: 4.3.12
[acquisition-board] ONI Driver: ft600 Version: 1.0.3
[acquisition-board] FTDI Driver version: 1.3.16
[acquisition-board] FTDI Library version: 1.3.16
[acquisition-board] Open Ephys ECP5-ONI FPGA open. Gateware version v1.5.3
[acquisition-board] Acquisition board ID = 256
[acquisition-board] Initializing ONI Acquisition Board...
[open-ephys] Creating processor with name: Record Node
[open-ephys] Creating processor with name: Channel Map
[open-ephys] Creating processor with name: LFP Viewer
[open-ephys] Creating processor with name: Bandpass Filter
[open-ephys] Creating processor with name: Spike Detector
[open-ephys] Creating processor with name: Audio Monitor
[open-ephys] Creating processor with name: Spike Viewer
[open-ephys] Beginning HTTP server on port 37497
[open-ephys] Checking for a newer version of the GUI...
[open-ephys] Starting acquisition
[open-ephys] Starting recording
[open-ephys] Stopping recording
[spike-detector] SpikeDetector detected 493 spikes in 562687 callbacks.
[open-ephys] Stopping acquisition

Please feel free to let me know if you have any thought.
I can do more bath recording for testing this but hopefully not animal recording.