>Number: 6490
>Category: user
>Synopsis: relayd check https - ssl read timeout
>Confidential: yes
>Severity: serious
>Priority: medium
>Responsible: bugs
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: unknown
>Arrival-Date: Fri Oct 15 19:10:01 GMT 2010
>Closed-Date:
>Last-Modified:
>Originator:
>Release:
>Organization:
>Environment:
System : OpenBSD 4.8
Details : OpenBSD 4.8-current (GENERIC) #175: Tue Oct 5 09:39:00
BST 2010
[email protected]:/usr/src/sys/arch/i386/compile/GENERIC
Architecture: OpenBSD.i386
Machine : i386
>Description:
using "check https" in relayd to check a service results in
something like 90%+ of checks failing:
8 host 192.168.41.2 9.21% down
9 host 192.168.41.3 9.21% down
occasionally we see
hce_notify_done: 192.168.41.2 (http code ok)
hce_notify_done: 192.168.41.3 (http code ok)
but most cheks fail like this:
hce_notify_done: 192.168.41.3 (ssl read timeout)
hce_notify_done: 192.168.41.2 (ssl read timeout)
>How-To-Repeat:
something like this (excerted from a longer configuration) should do it:
table <webmail> {
192.168.41.2
192.168.41.3
}
http protocol http-wm {
header change "Connection" to "close"
}
relay webmail-ssl {
listen on 195.95.187.10 port https
protocol http-wm
transparent forward to <webmail> check https "/" code 200 interface
vlan410
}
>Fix:
work-around: use 'check ssl' instead of 'check https "/" code 200'
>Release-Note:
>Audit-Trail:
>Unformatted:
