hard coded http timeout in haproxy_server role template
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack-Ansible |
Fix Released
|
Undecided
|
Chenjun Shen |
Bug Description
Dear all,
In our private cloud, we experienced some timeout issues in haproxy.
Some API requests took more than 50 seconds.
As we found out in current haproxy_server role haproxy.cfg.j2 template. The timeout are unfortunately hard coded, and there is no variable which you could use to override them.
Snippet of current haproxy.cfg.j2 template.
-------
defaults
log global
option dontlognull
option redispatch
retries 3
timeout client 50s
timeout connect 10s
timeout http-request {{ haproxy_
timeout server 50s
maxconn 4096
-------
I will implement some new variables which could be used to override timeout.
tags: | removed: haproxy timeout |
Changed in openstack-ansible: | |
assignee: | nobody → Chenjun Shen (cshen) |
status: | New → In Progress |
bugfix patch is here. /review. openstack. org/#/c/ 472200/
https:/