Correct html binding format in Dashboard2 - panel data freezes? Updated findings!

Hi, I have page with multiple Temperature and Humidity monitors. I have now transfered it to Dashboard 2 versions and having some intermittent ( typically after 16-24 hrs) freezing of those HTML elements (data is not updated correctly) on a always on tablet. Everything else like iFrame grafana embeddings and button / timer functions still work correctly though.
I did not have this issue in D1 version where I was using the ng-bind-html.

<div class="blocktitle orange1" <b>  T5-MAKUUHUONE</b></div>
<div class="{{msg.payload.Color2}}" ng-bind-html="msg.payload.TEMP5"></div>
<div class="blocktitle orange3" ng-bind-html="msg.payload.HUM5"</div>

As in the new D2 version this does not work I used this to get it working ( with additional text formatting) , but as stated this give me sometimes frozen data results.

  <div class="bg-grey-darken-1 text-center text-h6 ">
  <div class="bg-grey-darken-1 text-center text-h3 font-weight-bold">
  <div class="bg-grey-darken-1 text-right">

Output in to the Page looks like this, data inside red rectangle gets frozen from time to time.

After additional testing I found out that the issue is that as I am learning new CSS/ JavaScript / Vuetify coding If Make make "Wrong Enough" code and deploy it then the Pge gets "Hanged" even though the applied code is not for that specific page.

Bug or a feature ?

Applying refresh load on the table assumes normal operations of the page again ( untill I make a new "wrong enough" code deployment ).

Additional findings.

I created a html clock widget to understand when the page stops and on the system log file I found this when it stopped ? D2 is the name where all my pages reside.

10:11:03 SAMONE Node-RED[887]: 3 Mar 10:11:03 - [info] [ui-base:D2] Disconnected bFaAKBFddJvzUORBAAal due to ping timeout

Previously you said it hangs when have an error in the code. Is it now happening under a different situation or is this still caused by errors in the code?

Hi, Yes that was originally the assumption as I was trying (a lot ) to get the code working. But this time the freeze happened when I was not at the machine at all.

I tried to change the original text but I have lost possibility to edit it (again).

Aplogies for not beeing consistant in my ramblings :slight_smile:

Check how much memory used by dashboard. See how it grows... It may indicate problems if any

Looking on the time 10:10 no real changes on this. This used to be problem in D1 and was hoping D2 is not affected similarly (and data shows that to be true)

Did you say that it recovers if you refresh the page in the browser?

Is the browser running on the same machine that is running node red?

If you are able to, open the developer tools console in the browser and leave it open. See what it shows when it fails.

I wonder whether you are suffering from Page not updating when left in background on Android · Issue #444 · FlowFuse/node-red-dashboard · GitHub or Widget state not updated after network disconnection and recovery · Issue #620 · FlowFuse/node-red-dashboard · GitHub. I think they may be the same issue.

Yes, this page is run in separate Android tablet. And yes, manually refreshing the page it runs again for a while.

Thanks I will check the post you linked.

Memory leaks can be problem with Vue, vuetify, dashboard project or you create them by your coding. Blind blaming is worth nothing. But investigating is not trivial task. Specially for such kind of pretty large projects as dashboard + it's usage creates.

True, At this stage I always suspect my coding first :slight_smile:

I can confirm this is same bug that @Colin has identified prior. Should I flag somehow this in Github also ?

@Colin , unfortunately no devtools on the Android tablet so I could see inside what's happening.

You could add a comment on the github issue saying that you are seeing it too, and that is a pain. Hopefully it will get bumped up the priority list.

1 Like

Have bumped into the "Up Next" column, need to work out how to re-produce this issue quickly so I can develop.fix it.

Great, thanks, it will be good to have this working well.
#620 is fairly easy to reproduce and I am pretty sure it is the same issue (or at least part of it).

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.