Powerwall Dashboard data capture seems to stop #641
Unanswered
SeanNazareth
asked this question in
Q&A
Replies: 2 comments 3 replies
-
|
HI Sean, what do you see in the pypowerwall logs ( If so, I have an idea. We could add a health check to compose so that if pypowerwall is unable to pull data, it triggers a container restart. |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
I added healthchecks and a watchdog script for pypowerwall that may help. See #643 You should be able to install this by doing an upgrade. I always recommend doing a backup, but it should persist your existing historic monitoring data (influxdb). ./upgrade.sh |
Beta Was this translation helpful? Give feedback.
-
|
I'm seeing similar symptoms to Sean. PW3 running 25.18.1, accessed with an Opal router acting as a WiFi bridge. Dashboard is running on a VM elsewhere. So far when this has happened I can still ping the PW3. Here's two typical log sequence from Martin |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
Hi Jason, Though I can see from grafana that data collection stopped, I don't see any logs from that time. However, I restarted the pypowerwall on 6/19, and though the capture is currently running, I do see "failures" during the time that the capture has been running. Here is the start of the logs that are currently available through the first error. I see that there is a new version released today. Should I try to capture another "failure" where the data capture stops? Or re-setup with this new version with the health-check? Sean |
Beta Was this translation helpful? Give feedback.
-
Timeout. That is basically an indication that pypowerwall is waiting for the Powerwall to respond and gives up after 5s. It could be... Network: something in the stack is causing issues (congestion, WiFi signal, etc.) Host: if the host your using becomes too "busy", goes into power save mode or is otherwise overloaded, it can cause I/O issues Powerwall: something about it is not happy, most of the time it happens if the Powerwall thinks it is being attacked (DDOS) so it rate limits and blocks for some cool-down period. However unlikely, just make sure you don't have multiple things hitting your Powerwall at the same time. Tweaking pyPowerwallYou can try to tweak some of the proxy settings to better accommodate your network conditions. Edit the pypowerwall.env file and add these: PW_CACHE_EXPIRE=20 # How long cache will hold a response from Powerwall before fetching a new one - default = 5
PW_TIMEOUT=15 # How long pypowerwall will wait for the Powerwall to respond before a timeout - default = 5
PW_POOL_MAXSIZE=20 # Concurrent connection pool - default 15, and sometimes higher helps, sometimes lower helpsRestart the stack to apply: ./compose-dash.sh up -d |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hi,
I've set up an Odroid C4 (ARM) (Raspberry pi4-like SBC) to run the Powerwall dashboard (2x PW3) by connecting to the PW3 AP. For the most part it works. The odroid uses a USB-Wifi adapter to connect to the Powerwalls AP. The Odroid C4 is not too far away from the PW3, and is connected by ethernet to my home network. The issue is that it randomly stop capturing data. The ODROID C4 itself continues to be accessible, and the grafana dashboard continues to run, but it obviously doesn't show data for the period where the capture stops. I've seen it stop within hours of starting the docker stack, but it typically runs for 3-5 days before it stops. So far, I've restarted the data capture from the PW3 by rebooting the ODROID. I'm not sure (yet) if there is another method to recover/restart the capture of the production data.
Do you have any suggestions to debug this issue? Any suggestions to detect the problem and auto-restart the data logging?
Thanks,
Sean
Beta Was this translation helpful? Give feedback.
All reactions