I retried it about 15 times over the course of a few days.
That is strange. Was there anything in the logs - haven't touched the code in a while but if there was a typo that wouldn't let the code compile it will cause an error like that.
Think I found your issue, testing the fix right now. Missed that you were using the HTTP endpoint method when I looked before. As to 100% loss on a device that is up, only time I've that consistently is when I was testing against 8.8.8.8 (Google DNS).
Edit: 2.1.7 is now available to address the IP parameter passing bug
It appears that the 100% loss issue was either a cable or an unmanaged switch. For a test, I moved the hub directly to the router and all losses are gone, so I left it there. I hate coincidental failures. I just created a RM rule to repeat ping every xx seconds and that worked I but will grab the new version.
This update fixed the repeating actions (using the HTTP Endpoint).
Thanks for providing this. It's a great tool.
1 Like
Update to include the initialize option appears to work the way I want it to. Just restarted my hub and saw the schedule job come up for my device. Thanks again...
1 Like
I've been looking for a way to monitor when one my hubs is unavailable in order to automatically reboot it. Any reason I can't use this cool driver to fill that need? I'm picturing using RM, triggered when responseReady=TRUE to then evaluate Presence. If Not Present, then "Send Post to: http:/[hub-ip-address]/hub/restart", eh?
I'm going to try this. Any reason it won't work?
And am I essentially recreating this app inefficiently and with multiple hubs?
Any advice is greatly appreciated!
Edit: @thebearmay, I just tried the above. In RM, if you try to use a conditional with a custom attribute of responseReady, RM gives you the following error: "Not supported attribute type: BOOL". I assume that means it's a Boolean value that RM can't test for. If that's correct, I'm wondering if you can change the attribute format of responseReady to something else so that it's value can be evaluated by RM?
Interesting, never realized that an attribute couldn’t be a boolean. I’ll change it to a string.
2 Likes
yknow what's weird, there are other boolean fields that can be evaluated using RM, so I don't really know why it doesn't work in this case. Maybe someone more knowledgeable than me can comment.
Perhaps post a screenshot of how you were trying to do the comparison in the rule?
2.1.8 is out. Value is still true/false but is now a string.
1 Like
Wonder if they are actually ENUM with possibleValues set to true/false…
Wondered the same, so I tried it. No joy. Still stuck.
Edit: will try to version and revert.
Could there be a problem with HPM manifest? I updated via HPM no problem. It shows 2.1.8 is installed. But when I view the driver code itself, it still shows 2.1.7?
Just updated my production hub using HPM and got the right version, so not sure why you didn’t get the latest version. Might try doing a repair from HPM.
Thanks for making the change. Able to use RM now without a problem, thanks.
FWIW, I still get errors through HPM. Maybe it’s me:
Not sure about the HPM/github issue, as it would have been the same manifest and files for a repair as for the update, but have seen and gotten it before. Glad that it is working for you though.
Hey @thebearmay I was looking for the capability to notify me if one of my hubs goes offline - looks like this and a Notifier task will do the job - thanks!
I noticed there is no option to "Enable description text logging" ala a log entry when it changes to/from present. And debug logging also doesn't report the presence status. Was this by design?
I just didn't think about it - like everything else it started small and keeps growing. Would be fairly easy to add though, just want presence change logged or ??
Presence changed would make sense. Maybe ping started, stopped, success, failure as well?
1 Like
v2.1.10 is out with the Text Logging option
1 Like
