History: the storage strip

Next to each storage of a server is the “Pulse” history strip: when backups ran and whether they were saved in this storage. It's in the server list, in the “Storages” column, and on the server overview, in the “Storages and history” block.

How to read it

Each tick is the backup runs at that moment. Frequent backups (every 30 minutes) merge into a solid band; infrequent ones (every 6 hours) show as separate ticks.

If at one moment there were both saved and unsaved backups, the tick shows the worst. Right is now, left is the start of the period.

Periods

The switch next to the “History” heading: “today” (from 00:00 to 24:00), “7 d”, “30 d”, “90 d”, “all” (since the agent was connected). The panel remembers the chosen period. The longer the period, the denser the band: at 30 days one tick covers several hours.

Tooltip

The strip is split into invisible segments: “today” — by hour, 7 and 30 days — by day, 90 days — by week. Hover over the strip to see the tooltip for that segment. A click or tap pins it to the segment; the left and right arrows move to the neighboring segments. On a phone the tooltip is a card at the bottom of the screen.

What's in the tooltip:

• the storage and the segment: “Primary · 13:00–14:00”, “Secondary · yesterday”, “Primary · Oct 7 — Oct 13”;
• what was saved: “Backup saved · 13:48”, “All 4 backups saved · last 18:00”, “2 of 5 backups saved · last saved 12:00”, “None of 5 backups saved” or “No backups”;
• if not saved — “Last saved here: …” and “Cause: …”: the S3 response code or the restic line, as the agent saw them;
• if the segment has skipped backups — “Newer ones were ready while the previous backup was being sent — the latest ready one is sent”;
• windows that explain gaps, with times: “agent offline”, “server was rebooting”, “staging ran out of space”, “a component is not in the backup”, “no access to the storage: …”;
• a gray line “the panel was unavailable — the agent followed its plan; backup reports arrived when the panel was back”. This is not a backup problem.

The panel counts how many backups were expected by the plan in effect during that segment. If the plan changed, older segments are counted by the old profile.

Where the data comes from

The history is built from backup copies, not from the log. So it isn't lost after 90 days, as log lines are. A copy later removed by cleanup counts as saved on its day.

Time

Segments are in the time zone of the device where the panel is open: “today” is your today's hours, days are your days. If you are in a different time zone from the panel, times in notifications will differ: they use the panel's time zone.