As Denny said large state values that are Maps are very cpu intensive. It is because how the data is managed in the system. If i recall it is something like when it is read out of the database the whole value has to be converted to json. Then once in memory the json data has to be modified and then the whole data is converted back to the database format and then written again.
It is a fairly intensive process. If the backlog gets large enough it can take minutes to do a single post. You are best to process the backlog in as large of chunks as possible. Influxdb can process a lot more data on a single post then we are likely to send even with our largest backlog.
I currently am running influx logger with an influx 1.8DB. I have many devices logging data all is well, I want to add Hubinfo to the mix. Hubinfo requires advanced attributes which disables my other devices. Is it possible to run both ? I tried installing a second instance of influx logger named "influx logger2" which was an epic fail. I tried writing to two different DBs with no joy. Any help would be appreciated.
Running a second instance would be the correct approach. The InfluxDB app is designed for this. The instances do not need to write to different data bases, but can all write to the same database.
FWIW, I run several instances because I need different values for keep-alive and only post when a data value changes.
Hi,
Apologies if this had been asked before. I am trying to ship logs from Custom App ( not a Device) to influxdb using InfluxDB Logger. Is this even possible with this integration?
You will want to use the first link i provided you to capture the Hub Live Logging. That will send them to InfluxDB. Then use the Dashboard in that thread as well to view the records once imported.
FWIW, there is an experimental integration for remote syslog servers, but it doesn't work quite right yet. There is also the app that @mavrrick58 posted a link to which can send to Influx.
You might want to wait on that. V2.x has, by most accounts, a bizarre query language. V3 is in alpha now and has the same, or closer to, the same query language as V1.
Hi. Does anybody know what "Status: 408, Error: Connect to :8086: connect timed out, Headers: null, Data: null" actually means? I'm using Influx v2.x and I've had this configured for well over a year and was working fine up until a few days ago. The IP is correct and it's listening on port 8086. I updated the hub to v.2.4.0.151 about a week prior to when it stopped working, but that's all I have changed. The only other log messages I get other than the above is that it's dropping events based on the 5000 event max limit. Would really appreciate some help. Thanks!
You can look at documented InfluxDB response codes at this URL below
That URL doesn't include a response code of 408 though. So then you would want to look up standard HTTP response codes. A quick Google search indicates that is a "Timeout" as the message part of that error indicates. Check your influxDB server and make sure it is up and responding. If not you should probably start with restarting it.
I had an issue similar once - turned out a OS update reset my firewall settings. So when I was on that local machine it seemed normal - just no data getting in. If your influxdb app is showing locally but you can't get to it from external ... maybe?
Yeah, somehow my nic was changed to a Public profile (likely some OS update) where communication wasn't allowed. Flipping that back to private and re-instantiating the InfluxDB Logger app fixed it. Getting data again! Thanks all for your responses!
Well, I hadn't tried yet - just assumed it would fail, but now that your question suggests there's a chance, I'll put it on the list of things to do and report back.
