Project

General

Profile

Bug #589

Hermes cluster stats fails badly when it can't get local stats

Added by Anton Leonov almost 2 years ago. Updated almost 2 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Target version:
% Done:

100%

Estimated Time:
1.00
Deployment type:
AppScale - self deployed

Description

This was found when on one deployment haproxy process failed at master node.
When /cluster/stats/proxies endpoint was queried, master node attempted to get local stats and failed to get haproxy stats it failed badly and returned 500 error.
Instead, it should report this failure as following: {
"failures": {"<local_ip>": "Connection refused: ..."}
"stats": {"<active_lb_ip>": { ...<normal_stats>... }}
}

History

#1 Updated by Anton Leonov almost 2 years ago

  • Project changed from Engineering to AppScale
  • Deployment type set to AppScale - self deployed

#2 Updated by Anton Leonov almost 2 years ago

  • Status changed from Feedback to Closed
  • % Done changed from 80 to 100

Also available in: Atom PDF