Helpdesk story

Why can a branch suddenly not reach its file server after a firewall failover?

When one branch loses its file server right after a firewall failover, check whether the HA pair is in sync. In this story only Columbus is affected, the hubs are fine, and the unit that took over at 8:12 AM never received the 11:48 PM commit from the night before. The HA detail shows the pair out of sync, so Tier 1 can send it to the right engineer.

Watch the ticket get answered

Ticket #4495 · Tier 1 · 8:20 AM
Columbus can't reach the file server since this morning.

Only Columbus. The hubs are fine.

Only one site is affected and the hubs are fine. So it's local, and Tier 1 can say so straight away.

Columbus: HA pair failed over at 8:12Passive unit now active · config not in sync
  • Columbus

The new active unit is missing last night's change.

Columbus-850a ↔ Columbus-850bHA failover 8:12 AM · now active: Columbus-850b

Columbus-850a

was active

Last commit11:48 PM by d.patel

Columbus-850b

now active

Last commitdoes not have the 11:48 PM commit

Config not in sync with peer · failover at 8:12 AM

Escalated with the cause attached: the pair is out of sync.

Step by step

Ticket #4495 · Tier 1 · 8:20 AM: “Columbus can't reach the file server since this morning.”

Sites and hubs: Only Columbus. The hubs are fine.

Only Columbus is affected and the hubs are fine, so the problem is local. The HA pair there failed over at 8:12 AM.

  • Sites affected: Columbus.
  • Columbus: HA pair failed over at 8:12. Passive unit now active · config not in sync.

HA detail: The new active unit is missing last night's change.

The HA detail shows Columbus-850b now active, and not in sync with its peer. It does not have the 11:48 PM commit that d.patel made on Columbus-850a.

  • Columbus-850a ↔ Columbus-850b
  • Failover at 8:12 AM; now active: Columbus-850b.
  • Config in sync with peer: no.
  • Columbus-850a (was active), last commit: 11:48 PM by d.patel.
  • Columbus-850b (now active), last commit: does not have the 11:48 PM commit.

The answer: Escalated with the cause attached: the pair is out of sync.

How Firewall Fleet Health spots this

  • Shows which sites are affected and whether the hubs are fine, so a local problem is obvious.
  • Shows each HA pair with the time of the failover and which unit is active now.
  • Shows whether the active unit is in sync with its peer, and the last commit on each unit.
  • Keeps 18 months of firewall events and config changes, so the failover and the earlier commit are both on record.

Common questions

How do I know if an HA failover left a pair out of sync?

Open the HA detail for the site. Firewall Fleet Health shows the failover time, the active unit, whether the config is in sync with its peer, and the last commit on each unit.

How can Tier 1 tell a local problem from a fleet-wide one?

Check the sites and hubs on one screen. If a single site is affected and the hubs are healthy, the problem is local and can go straight to the engineer for that site.

See your own fleet this way

Explore the demo fleet yourself.

Schedule a demo Explore the live demo (opens in a new tab)