mqtt_vacuum_camera: 2025.6.0b0 causes excessive network use
Checklist
- I have updated the integration to the latest version available
- I have checked if the problem is already reported
The problem
Updating to the 2025.06.0b0 version causes excessive network use, affecting system ping times:
Reverting to version 2025.05.0b7 causes the issue to disappear.
What version of an integration has described problem?
2025.06.0b0
What was the last working version of an integration?
2025.05.0b7
What vacuum model do you have problems with?
Roborock S5
Please firmware installed on your Vacuum.
No response
What type of platform you use?
Intel NUC (or generic x86_64)
What version of Home Assistant do you use?
core-2025.07.0
What type of installation are you running?
Home Assistant OS
Logs or Errors shown in the HA snapshots (please enable the Debug Mode) text will be auto formatted to code.
Function, that in your opinion is creating the issue.
None
Additional information
No response
About this issue
- Original URL
- State: open
- Created a year ago
- Comments: 16 (12 by maintainers)
Commits related to this issue
- correction for issue #366 Signed-off-by: Sandro Cantarella <sandro@cc497374-b62a-4cdb-ac2d-85600f69439b.fritz.box> — committed to sca075/mqtt_vacuum_camera by sca075 a year ago
Unfortunately it seems to be worse:
also:
Might have spoken too soon. This is killing my network when the vacuum is running. Remotely connecting to my home assistant instance is very slow. It may be affecting other network integrations too, Vacuum started at 11:00
I’m running HAOS on an i7 NUC (with Mosquitto MQTT add-on).
I’m not seeing a huge increase in network traffic using the system monitor integration, but am not sure if that includes the mqtt addon:
@tomlut removed all images from the coordinators data… we got this as result the last peak and one of the first are reboot data.
@tomlut okay I think I got it… HA coordinators use the event buss to exchange data, I didn’t know that… the problem was that updating the coordinator with poling and transferring images via the coordinator data was creating too much load on the network as you luckily reported. So, I now change from polling to on-demand updates and moved to event the client reception of the maps data… also store the map image that can be now retrieved on demand from the camera. Will monitor for a couple this changes and possibly we are safe to release finally.
@tomlut I think I got were is the problem and I’m very close to fix this… those data are from the modifications I’m currently testing some of the peaks are because of the tests.