View Issue Details

IDProjectCategoryView StatusLast Update
0006191CentOS-6squidpublic2013-01-14 16:59
ReporterRolandS 
PrioritynormalSeveritycrashReproducibilityalways
Status newResolutionopen 
PlatformIntel Xeon E5530 (VM)OSCentos 6.3 x86_64OS Version6.3 Final
Product Version6.3 
Target VersionFixed in Version 
Summary0006191: Using squidclient tool makes squid crash
DescriptionThis bug report follows this forum entry : https://www.centos.org/modules/newbb/viewtopic.php?topic_id=40914

On the following CentOS installation:
- CentOS 6.3 with latest "yum upgrade"
- Official kernel 2.6.32-279.19.1.el6.x86_64
- Latest Squid package squid-3.1.10-9.el6_3.x86_64 from official repository
- SELinux enforced

The configuration file allows the use of cache_object, and squid works very well on this server as long as I just ask him to relay/cache HTTP requests.
But the use of such a command like "squidclient mgr:info" (on the host itself) makes it crash instantly, with a mail to root from auditd that follows.
Steps To ReproduceHaving squidclient allowed :
- acl manager proto cache_object
- http_access allow manager localhost

Simply launch :
- squidclient mgr:info

Results on :
- squid crashes
- squidclient returns nothing to the console and exits
- auditd sends a mail about the crash to root mailbox
Additional InformationTo know for sure that it is a "Aborted (core dumped)" issue, I stopped the service then launched squid in interactive mode with this command :
- squid -N -X -f /etc/squid/squid.conf

As an alternative to squiclient I also used netcat :
- nc 127.0.0.1 3128
Then on the standard input (with carriage return) :
GET cache_object://localhost/info HTTP/1.0

With the same result.
TagsNo tags attached.

Activities

There are no notes attached to this issue.

Issue History

Date Modified Username Field Change
2013-01-14 16:59 RolandS New Issue