Swift: a server process is terminated ungracefully when stopping its container
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
In Progress
|
Medium
|
Takashi Kajinami |
Bug Description
Description
===========
Currently SIGTERM is widely used to stop processes inside containers.
However, sending SIGTERM to a swift-*-server process terminates the process ungracefully, and this can cause interruption of on-going API requests.
For swift*-server process, we should use SIGHUP so that the process stops gracefully after it completes all on-going API requests.
Steps to reproduce
==================
* Stop swift_proxy container
$ sudo systemctl stop tripleo_swift_proxy
* Check swift.log
Expected result
===============
swift-proxy-server process receives SIGHUP and stops gracefully
Actual result
=============
swift-proxy-server process receives SIGTERM and stops immediately
Jul 18 23:39:25 controller-0 proxy-server[6828]: SIGTERM received
Jul 18 23:39:25 controller-0 proxy-server[6828]: Exited (7)
Environment
===========
The issue was initially found in stable/train deployment
Logs & Configs
==============
N/A
Changed in tripleo: | |
assignee: | nobody → Takashi Kajinami (kajinamit) |
Changed in tripleo: | |
importance: | Undecided → Medium |
Fix proposed to branch: master /review. opendev. org/c/openstack /tripleo- heat-templates/ +/801197
Review: https:/