subpage with 4 charts fast at first and then slows down consuming more and more CPU

[christf]

Current Behavior

I recently migrated from the dashboard 1 to node-red-dashboard-2 with node-red 4.0.9.

On one of my pages, I have 4 bar charts that get pushed data that looks like this:

{"payload":[{"Reihe":"Verbrauch","Medium":"Netz","Wert":0},{"Reihe":"Verbrauch","Medium":"Solar","Wert":638},{"Reihe":"Verbrauch","Medium":"Akku","Wert":0},{"Reihe":"Erzeugung","Medium":"Netz","Wert":15},{"Reihe":"Erzeugung","Medium":"Solar","Wert":638},{"Reihe":"Erzeugung","Medium":"Akku","Wert":177},{"Reihe":"Verbraucher","Medium":"Heizung Sophie","Wert":443.4},{"Reihe":"Verbraucher","Medium":"Heizung Bad","Wert":0},{"Reihe":"Verbraucher","Medium":"Heizung Jojo","Wert":0},{"Reihe":"Verbraucher","Medium":"Heizung Lydia","Wert":0},{"Reihe":"Verbraucher","Medium":"Heizung Wohnzimmer","Wert":0},{"Reihe":"Verbraucher","Medium":"Waschmaschine","Wert":0},{"Reihe":"Verbraucher","Medium":"Brunnenpumpe","Wert":0}],"topic":"Messwerte2","_msgid":"8ea04fd3362b7753","headers":{"content-type":"application/json;","cache-control":"no-store, no-cache, must-revalidate, post-check=0, pre-check=0, max-age=0","content-length":"713","date":"Mon, 16 Mar 2026 14:04:02 GMT","server":"lighttpd/1.4.54","x-node-red-request-node":"0908632b"},"statusCode":200,"responseUrl":"https://sma-wr/dyn/getValues.json?sid=xxxxxxxx","redirectList":[],"retry":0,"sid":"xxxxxxxxx","freie_solarenergie":192,"P_solarbilanz":192,"ui_update":{"chartOptions":{"xAxis":{"type":"category","showGrid":false,"splitLine":{"show":false}},"yAxis":{"max":5000}}}}

this worked great before the migration. Now I am observing that somehow the page slows down.
I have tried limiting the frequency, which is possible to do for 3 of the 4 bar charts. This seems to delay the problem but eventually the browser tab displaying the page consumes 100% CPU and the only way to recover is to stop and restart it.

Expected Behavior

this should not slow down the page nor delay updates.

Steps To Reproduce

• deploy node-red 4.0.9
• use the new dashboard 2
• deploy a chart element
• feed it data as above at 2s interval

It seems to be a bit better with the msg.headers, msg.responseURL and msg.sid fields removed.

This is the node-config I have in place:

Environment

• Dashboard version: 2
• Node-RED version: 4.0.9
• this is running with the docker image docker.io/nodered/node-red:4.0.9

Have you provided an initial effort estimate for this issue?

[colinl]

What version of dashboard 2 are you using? Check in Manage Palette that you are using the latest version.

[christf]

this is 1.30.2 as per the manage palette dialog

[colinl]

Could you Export a simple flow (Inject node + chart) please and paste here (as text). That will make sure I have exactly the same options as you in the chart.
What browser are you using? Do you see the same problem in a different browser?

[colinl]

First though, remove the ui_update section of the message. It is only necessary to send that once, so use an inject once at startup for that, or something similar. There is a known issue that repeatedly sending the options can clog things up.

[christf]

when removing ui_update, the symptom is gone. Thank you for this hint. :)
Since this bar chart requires changing scale during runtime, my current workaround is to keep the state of the chart in a variable and only send ui_update, when changes are required. Still, it is a regression when migrating from the ui to dashboard.

[colinl]

There is already an open issue to get this sorted. #2012