MQTT Transmission History Settings and Limits

I am curious what is the better option for the history store type for the Transmission module. I have Panel PCs; the only thing it’s doing is running Ignition (Visualization, Computer, and IIoT modules).

The number of tags, approx: 150 to 200, is the number of tags being transmitted. I need to know how much storage (In-Memory or Disk-Backed) I need. Do I need to attach an additional hard drive or not?

Let’s use a downtime of 1 week to start.

Take a look at this doc: Determining the settings for an MQTT Transmission History Store - MQTT Modules for Ignition 8.x - Confluence

In that video on that page you linked, what units is the Memory Used tag in? MB?

It is in just bytes.

I get an error when setting the Device Tag Capacity to 10000000000 (10,000,000,000). The error message is below.

The value of ‘Device Tag Capacity’ is not a valid Integer.

I let the edge system (56 tags changing once a second) run for 24 hours without a connection to the broker and it used 36% of the metrics used and didn’t even use 1 GB (I approx. about 0.58 GB in 3 days, which would max out).

Now my Edge systems will be running on a Panel PC with 256 SSD and I need it to buffer for longer than 3 days (preferably 2 weeks). With it not even using 0.58 GB in 3 days I would think it could handle that but if the Device Tag Capacity is reached in 3 days, then I will start losing records that are 3 days old.

Should I be increasing the Device Tag Capacity or Edge Node Tag Capacity to achieve this?

Which you increase depends on whether the tags live under the ‘Edge Node Level’ or the ‘Device Level’. We’re currently working on an improvement in Transmission which will use a time based expiration rather than a count based expiration (i.e. two weeks instead of 100,000 data points). If all goes well this will be in the 4.0.18 version of the Transmission module.

For specifics on the edge node vs device levels, see this: MQTT Transmission Transmitters and Tag Trees - MQTT Modules for Ignition 8.x - Confluence