Bug 754916

Summary: netdump client always access /var/crash even though netdump server specify custom dump dir
Product: Red Hat Enterprise Linux 4 Reporter: masanari iida <masanari_iida>
Component: netdumpAssignee: Neil Horman <nhorman>
Status: CLOSED WONTFIX QA Contact: qe-baseos-daemons
Severity: medium Docs Contact:
Priority: unspecified    
Version: 4.8   
Target Milestone: rc   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-11-18 11:56:14 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description masanari iida 2011-11-18 08:18:15 UTC
Description of problem:
netdump client always try to create magic file into /var/crash/magic,
even though netdump server specify custom dump dir.

Version-Release number
netdump-0.7.16-15

How reproducible:
Always, if netdump server specify custom dump directory.


Steps to Reproduce:
Follow the KB (https://access.redhat.com/kb/docs/DOC-63704),
create a netdump server with custom dump directory.

1. Configure netdump server.  Specify --dumpdir in /etc/sysconfig/netdump-serer.

ex.
SERVER_OPTS="--dumpdir /test/crash"

2. Start netdump server .

3. Start netdump client

4. Crash netdump client.

  
Actual results:
netdump client created magic file under /var/crash/magic/IP 
everytime netdump client started.

Then when the system crash by SYSRQ key, the netdump client sent
syslog file to /test/crash/IP/log.
But no vmcore file or vmcore-incomplete file were saved.

magic file under /test/crash/magic/ip was not updated all the time.

I see "Got invalid magic number (ff6b28649e61e512) from client IP " 
in messages. 

Expected results:
The vmcore file is saved into custom dump directory.


Additional info:
The root cause is, path to storing magic file is FIXED value.

/etc/init.d/netdump 

start ()
{
        *)
    	    ssh -x -i /etc/sysconfig/netdump_id_dsa netdump@$NETDUMPADDR echo "$MAGIC2$MAGIC1" \> /var/crash/magic/$LOCALADDR
}

Modify the path to /test/crash/magic/$LOCALADDR, and re-start the netdump
client again.  Then crash the system.  The vmcore was generated succesfully.

Also, propagete() need to be fixed.

Comment 1 Neil Horman 2011-11-18 11:56:14 UTC
RHEL4.8 entered its Maintenance phase at the start of this year, so this won't be fixed.  To work around the problem you can disable the exchange of magic numbers by using the --secure 0 flag on the netdump server