I've just upgraded my test server to 4.0.3 and there is a problem with (I think) hobbitd_larrd. I was previously using hobbit 4.0.2 with no problem. The server is running Centos 4.0 (Redhat AS 4.0 equiv) The various config files are as installed by the hobbit install process, only the alerts file is changed. I'm providing sendmail and mailq data via the BB scripts.
In the larrd-status.log I get:
2005-05-22 03:44:49 Tried to down BOARDBUSY: Invalid argument 2005-05-22 15:50:53 Gave up waiting for GOCLIENT to go low. 2005-05-22 16:37:50 Tried to down BOARDBUSY: Invalid argument 2005-05-23 04:18:51 Gave up waiting for GOCLIENT to go low. 2005-05-23 16:45:29 Tried to down BOARDBUSY: Invalid argument 2005-05-23 17:08:36 Tried to down BOARDBUSY: Invalid argument 2005-05-23 17:10:07 Tried to down BOARDBUSY: Invalid argument
In the larrd-status.log I get: 2005-05-23 17:10:07 Tried to down BOARDBUSY: Invalid argument *** glibc detected *** corrupted double-linked list: 0x00479838 *** 2005-05-23 17:11:45 Worker process died with exit code 134, terminating *** glibc detected *** corrupted double-linked list: 0x00479838 *** 2005-05-23 17:12:54 Worker process died with exit code 134, terminating
ANy suggestions? I was planning to put the server in as a production server tomorrow morning :-(
Regards Geoff
On Mon, May 23, 2005 at 05:21:17PM +1000, Geoff Steer wrote:
I've just upgraded my test server to 4.0.3 and there is a problem with (I think) hobbitd_larrd. I was previously using hobbit 4.0.2 with no problem. The server is running Centos 4.0 (Redhat AS 4.0 equiv)
In the larrd-status.log I get: 2005-05-23 17:10:07 Tried to down BOARDBUSY: Invalid argument *** glibc detected *** corrupted double-linked list: 0x00479838 *** 2005-05-23 17:11:45 Worker process died with exit code 134, terminating
There ought to be a core-file from this somewhere - in ~hobbit/server/tmp/ probably. Could you check for this and send me the traceback (see http://www.hswn.dk/hobbit/help/known-issues.html#bugreport if you're not familiar with how to do this) ?
-- Henrik Storner
On Mon, May 23, 2005 at 05:21:17PM +1000, Geoff Steer wrote:
I've just upgraded my test server to 4.0.3 and there is a problem with (I think) hobbitd_larrd.
In the larrd-status.log I get: *** glibc detected *** corrupted double-linked list: 0x00479838 *** 2005-05-23 17:11:45 Worker process died with exit code 134, terminating
Geoff provided some more info about this, and it appears to be a problem with the RRDtool library - he was using the newer 1.2.7 version. Downgrading to RRDtool 1.0.50 solved the problem.
Henrik
participants (2)
-
gsteer＠firstwave.com.au
-
henrik＠hswn.dk