[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference vaxaxp::vmsnotes

Title:	VAX and Alpha VMS
Notice:	This is a new VMSnotes, please read note 2.1
Moderator:	VAXAXP::BERNARDO

Created:	Wed Jan 22 1997
Last Modified:	Fri Jun 06 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	703
Total number of notes:	3722

619.0. "Vax V6.2:shadowing problem" by TOSSUB::PEGNA () Tue May 20 1997 06:39


	Hi,

	My customer is experiencing a file corruption problem many time in 
	less than 1 mounth in the same shadowset (different file)
	
	The enviroment is:  2 vax 3500 running openvms v6.2 
	using cluster and shadow with patch  VAXSHAD05_062 installed.




	node A			exp. box		node B
				 r215f
	---------		---------		---------
        |dua0   |		|dia4   |		|dua1	|
        |dua2   |		|dia5   |		|dua3	|
        |       |		|	|		|	|
        |	|	dssi	|	|	dssi	|	|
        |kzqsa  |---------------|	|---------------|kzqsa	|
        |       |		|	|		|	|
        ---------		---------		---------
            |			    			    |			
	    |						    |
	----|-----------------------------------------------|--------
				ethernet



	$1$dia5 (rf72) is quorum disk	

	$1$dia4 (rf72) is member of shadowset dsa45 (only one member) 

	$1$dua0 (ra70) is system disk node A 

	$1$dua1 (ra70) is system disk node B

	$1$dua2 (ra70) and $1$dua3 (ra70) are member of shadowset dsa23

	
	The application's customer ,in basic, updates the files on dsa23
	and it seems to work fine.The errorlog.sys doesn't show any error.
	
	Random, a file on dsa23 is right if the user works from node A,
	is bad from node B.
	DIR/SIZE from node A show a different size of file  from node B
	After dismount of dsa23 I do a mount/over of dua2 and dua3
	and check the file :the size is different ,also the dump/file. 
	
	
	This problem occur after upgrade from vms v5.5-2 to v6.2
		

	Thanks,
	Beppe

T.R	Title	User	Personal Name	Date	Lines
619.1		UTRTSC::utojvdbu1.uto.dec.com::JurVanDerBurg	Change mode to Panic!	`Tue May 20 1997 10:22`	5
	Set NISCS_PORT_SERV to 3 on both nodes to rule out that the corruption comes from network components. What kind of corruption? Complete blocks? Just a couple of bytes? Jur.
619.2		TOSSUB::PEGNA		`Thu May 22 1997 09:09`	12
	I set NISCS_PORT_SERV to 3 on both nodes. The files mov002.du3 and mov002.du2 are in ITTO01$DKB700:[colombo] on node ITTO01 (46.317) MOV002.DU3 is right (from $1$dua3 member of dsa23 shadowset) MOV002.DU2 is bad (from $1$dua2 member of dsa23 shadowset) Beppe.
619.3		UTRTSC::jgoras-197-2-3.jgo.dec.com::JurVanDerBurg	Change mode to Panic!	`Thu May 22 1997 10:59`	11
	> I set NISCS_PORT_SERV to 3 on both nodes. Did the corruption occur since then? A couple of blocks in the file look similar but different. This can also be an application problem. For example, an application writes to a shadowset and before the i/o completes it modifies it's own databuffer. How do they write this file? Is it an async write? Jur.
619.4	VAXSYS01_062 installed?	JARETH::NORWICH		`Wed May 28 1997 13:31`	6
	Is VAXSYS01_062 installed? Is the data on the disks incorrect or is it only bad "in memory"? Phil
619.5		TOSSUB::PEGNA		`Tue Jun 03 1997 12:08`	21
	> RE .3 > Did the corruption occur since then? Yes, the corruption occur again. > RE .4 > Is the data on the disks incorrect or is it only bad "in memory"? The data is incorrect on the disk. > Is VAXSYS01_062 installed? I installed it yesterday (jun 02) . Beppe
619.6	And after you installed the patch?	JARETH::NORWICH		`Tue Jun 03 1997 20:09`	10
	Has the corruption occurred since you have installed VAXSYS01_062? If the answer is yes, then I would suggest you IPMT it. You will most likely be asked for dumps of the File(s) in question. How you noticed the corruption. Any significant events envolving the device(s), etc. Phil