fleet
fleet copied to clipboard
fleetdm
Delay in updating policy pass/fail counts
Fleet version: 4.49.3
Web browser and operating system: Chrome
💥 Actual behavior
A policy count shows 2 failing hosts ("macOS - Check if latest version"), but when I click the link, I see 25 hosts that are listed as showing. If we know the 25 failing hosts, we should update the policy count. It did eventually update.
🧑💻 Steps to reproduce
- Create a new "macOS - Check if latest version" policy.
- Wait for initial results.
- Click on the host count and see that the filtered list of hosts does not match the host count.
🕯️ More info (optional)
N/A
@lukeheath by design the cron job that updates the total number runs every 1H. We could implement some UI coding that will collect the current status (Or do the same in BE) but that will be a feature request.
If still required, we can make it a story. In such case we'll need product to describe how to solve it. (e.g. UI will count current failing hosts, or more frequent cron jobs etc)
the cron job that updates the total number runs every 1H.
@lukeheath here's the issue where we made this change: #13574
I think 1 hr is for customers with a large number of policies and hosts. We could crank it down in dogfood.
But, I don't think this job's frequency is configurable. @mostlikelee please correct me if I'm wrong.
Maybe that's the feature request: making this job configurable.
But, I don't think this job's frequency is configurable. @mostlikelee please correct me if I'm wrong.
That's correct. We could also consider defaulting smaller environments to calculate counts on demand. The original issue is that the request/page load would take a very long time with large amounts of hosts and policies.
@lukeheath we added this tooltip a few months ago, does this solve it or do you think that info needs to be more prominent?
Removing from feature fest for now. @lukeheath please feel free to bring it back.
Policy counts in sync, Like leaves aligned in the breeze. Clarity takes root.
