[Orca-users] orcallator segfaults

[Buhler Justin]

Actually the problem exists in two places - 
 
old versions of SE < 3.3.1 
 
and the RAW_Disk code in orca as Dmitry already mention.
 
If you upgrade the SE and have the latest orcalletor.cfg the problem should go away. If that is not an option then comment out the RAW_DISK 1.
 
Justin
 

	-----Original Message----- 
	From: orca-users-bounces+jpbuhler=athens2004.com at orcaware.com on behalf of David Devault 
	Sent: Wed 8/25/2004 8:39 PM 
	To: Tony Howat; orca-users at orcaware.com 
	Cc: 
	Subject: RE: [Orca-users] orcallator segfaults
	
	

	I bet this is a Solaris problem and not an Orca problem.
	
	I don't know if this is what the problem is but you should check these things anyways.  Making these type of changes can severely hurt your system. proceed at your own risk.. If you do not know what this information is that I'm referring to - check, check and check again.  All of these files and configurations are very important for the way that Solaris works. 
	
	When the server boots the World Wide Names are printed to the console and are found in dmesg output.  These WWN's and SCSI ID's need to be added to your fiber channel card configuration file.  Usually located in /kernel/drv/somfilename.conf, I think you can make your device names shorter by adding the WWN's and SCSI ID's device instances and targets into the files: /kernel/drv/sd.conf  AND your FC card configuration file.
	
	Check with your FC card vendor documentation and verify the exact configurations.
	
	Also verify with Sun documentation that the configurations in sd.conf are correct and that you do not need to add any other configurations for the SAN.  There may be something missing in the /etc/path_to_inst file after you make those changes and it may need to be rebuilt.
	
	Before doing any changes back everything up and store it somewhere else.  You will need to reboot with a /reconfigure to commit the changes.  If you can't make the changes bscause the system can not be brought down... then there might be something that can be changed in Orca...
	
	David
	
	-----Original Message-----
	From: Tony Howat [mailto:t.howat at arts.ac.uk]
	Sent: Wednesday, August 25, 2004 6:16 AM
	To: orca-users at orcaware.com
	Subject: [Orca-users] orcallator segfaults
	
	
	orcallator is segfaulting when I try to start it on a Solaris 9 R280R
	connected to a FC SAN....
	
	/opt/RICHPse/bin/se -DWATCH_OS -DWATCH_WEB -d /usr/local/lib/orcallator.se
	
	[lots of debug]
	if (strncmp(first_name<sd6>, 51Dd0<md103>, first_len<3>) == <0>)
	if (strcmp(RAW_disk[3].short_name[3]<5>, 51Dd0<md103>) == <0>)
	j++;
	strcpy(51Dd0<md103>, GLOBAL_disk_info[43].short_name<md102>)
	if (j<43> > <0>)
	if (strncmp(first_name<sd6>, 51Dd0<md102>, first_len<3>) == <0>)
	if (strcmp(RAW_disk[3].short_name[3]<5>, 51Dd0<md102>) == <0>)
	j++;
	strcpy(51Dd0<md102>, GLOBAL_disk_info[44].short_name<(nil)>)
	Segmentation Fault
	
	I saw a posting
	http://www.orcaware.com/pipermail/orca-users/2003-December/003482.html
	
	that seems to report my problem. Ed claims that the problem is long/odd
	device names, citing his fiber devices with their WWNs embedded in them.
	This system also has FC devices connected, and has /dev/dsk containing
	things like :
	
	  ../../devices/pci at 8,700000/SUNW,qlc at 2/fp at 0,0/ssd at w200800a0b813180d,3:a
	lrwxrwxrwx   1 root     root          70 Jul 26 13:16
	c3t200800A0B813180Dd3s1 ->
	
	and:
	
	../../devices/scsi_vhci/ssd at g600a0b80001323b80000000040f2051d:b
	lrwxrwxrwx   1 root     root          63 Jul 29 10:13
	c5t600A0B80001323B80000000
	
	Has this problem been corrected?
	
	I am using orca 0.27
	
	--
	Tony Howat
	UNIX Network Administrator
	University Of The Arts London
	
	_______________________________________________
	Orca-users mailing list
	Orca-users at orcaware.com
	http://www.orcaware.com/mailman/listinfo/orca-users
	
	_______________________________________________
	Orca-users mailing list
	Orca-users at orcaware.com
	http://www.orcaware.com/mailman/listinfo/orca-users