Project

General

Profile

Bug #45876

cephadm: handle port conflicts gracefully

Added by Sebastian Wagner 4 months ago. Updated 4 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
cephadm/scheduler
Target version:
-
% Done:

0%

Source:
Tags:
ux
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature:

Description

INFO:cephadm:Verifying port 9100 ...

 WARNING:cephadm:Cannot bind to IP 0.0.0.0 port 9100: [Errno 98] Address already in use

 ERROR: TCP Port(s) '9100' required for node-exporter is already in use

 Traceback (most recent call last):

File "/usr/share/ceph/mgr/cephadm/module.py", line 1638, in _run_cephadm

code, '\n'.join(err)))

 RuntimeError: cephadm exited with an error code: 1, stderr:INFO:cephadm:Deploying daemon node-exporter.ceph-mon ...

 INFO:cephadm:Verifying port 9100 ...

 WARNING:cephadm:Cannot bind to IP 0.0.0.0 port 9100: [Errno 98] Address already in use

 ERROR: TCP Port(s) '9100' required for node-exporter is already in use

 2020-05-15T13:33:46.966159+0000 mgr.ceph-mgr.dixgvy (mgr.14161) 678 : cephadm [WRN] Failed to apply node-exporter spec ServiceSpec(

{'placement': PlacementSpec(host_pattern='*'), 'service_type': 'node-exporter', 'service_id': None, 'unmanaged': False}

): cephadm exited with an error code: 1, stderr:INFO:cephadm:Deploying daemon node-exporter.ceph-mon ...

 INFO:cephadm:Verifying port 9100 ...

 WARNING:cephadm:Cannot bind to IP 0.0.0.0 port 9100: [Errno 98] Address already in use

 ERROR: TCP Port(s) '9100' required for node-exporter is already in use

Important bits are:

  • We already know which services want which ports.
  • we can easily prevent port conflicts for known daemons.
  • open Q: how to handle unknown daemons (i.e. pre-existing node expoter)

Related issues

Related to Orchestrator - Feature #45652: cephadm: Allow user to select monitoring stack ports New
Related to Orchestrator - Feature #45565: cephadm: A service should provide information about itself New
Related to Orchestrator - Bug #45093: cephadm: mgrs transiently getting co-located (one node gets two when only one was asked for) Fix Under Review

History

#1 Updated by Sebastian Wagner 4 months ago

  • Related to Feature #45652: cephadm: Allow user to select monitoring stack ports added

#2 Updated by Joshua Schmid 4 months ago

  • Subject changed from cephadm: handle port conflicts gacefully to cephadm: handle port conflicts gracefully

#3 Updated by Sebastian Wagner 4 months ago

  • Description updated (diff)

#4 Updated by Sebastian Wagner 4 months ago

  • Related to Feature #45565: cephadm: A service should provide information about itself added

#5 Updated by Sebastian Wagner 3 months ago

  • Related to Bug #45093: cephadm: mgrs transiently getting co-located (one node gets two when only one was asked for) added

Also available in: Atom PDF