DEGRADED MODE — CACHED ENTITLEMENTS · SENTRIS-DB01 UNREACHABLE · People & Places duty desk
Modify ↓

Opened 13 years ago

Last modified 13 years ago

#23 RESOLVED fault

Telensa CMS intermittently cannot resolve node-registry hostname

Reported by: People & Places Owned by: D. Pike (Node Control)
Priority: P2 Milestone:
Component: DNS / Resolution Version:
Keywords: Cc:
Desk: Estate & Systems Legacy ref: DEVCON-1140

Description

The CMS periodically fails to reach the node registry and reports the whole estate unreachable for 30–120s, then recovers. Escalated in a panic as a total outage. Nodes were fine throughout; only the CMS's view of them dropped.

Attachments (0)

Change History (3)

comment:1 by S. Rao, 13 years ago

Estate 'down' for two minutes then back. Nodes never dropped association. It's not the mesh.

comment:2 by D. Pike, 13 years ago

It's DNS. The CMS caches the registry A-record on a 60-second TTL against a secondary resolver that keeps timing out.

comment:3 by D. Pike, 13 years ago

Pinned it to the primary resolver, bumped the TTL, killed the flapping secondary. Six days clean. Told you. Resolved.

Modify Ticket

Change Properties
Set your email in Preferences
Action
as RESOLVED The owner will remain D. Pike (Node Control).

Add Comment
E-mail address and name can be saved in the Preferences .
Attachments ↑ Description ↑
 
Note: See TracTickets for help on using tickets.