Read events on one of your machines.
Accounts server keeps its own disk warning, address change, and hello. Reception PC's reboot is not on this tab.
Open History › events
Open Accounts server, then history › events. The header is Melbourne / Accounts server, demo-dc1 · linux, version 0.9.37, offline · last seen 5d, with chips server, site/Melbourne, and role/DC. inbox · 1 and discuss are on the header. There is no control. History's tabs are events and trends. Audit is not a tab. The bar says last 7 days, all events, any severity. The count is 3. Newest first: 21/09 04:00 disk_high, warn, watchdog, disk / at 92%; 21/09 02:10 external_ip_changed, warn, agent, external IP changed 203.0.113.40 → 203.0.113.50; 20/09 22:10 hello, info, agent, connected v0.9.37. Reception PC's reboot is on the company list, not on this machine. This shot does not open a row.
Keep one event name
Set the event menu to disk_high. The address change and the hello drop out. The disk line stays, 21/09 04:00, warn, disk / at 92%. The count is 1 of 3. The other two lines are still stored. Set the menu back to all events and they return. Everything in the range menu is the same history, capped at 500 rows.
Every machine you can see is Read events on your machines. One name on that page is Show only one kind of event on your machines. Warnings only, on this tab, are Show only warnings on one of your machines. The last day is Show one of your machines' events from the last day. A word in the summary is Search events on one of your machines. What your provider sees, including audit, is Read one machine's events.