[Veritas-bu] EDT-AdvanTAPE driver problems with AIX 4.3.3 & N etBackup

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C202BB.01E240A0
Content-Type: text/plain;
        charset="iso-8859-1"

This is a known problem with Netbackup irrespective of tape drive in use.
 
See http://seer.support.veritas.com//docs/245505.htm
<http://seer.support.veritas.com//docs/245505.htm> 
 
1. You want to ensure scsi reserve/release is enabled on all media servers.
2. You want to ensure there is no device/tape monitoring software running on
the media servers
    that is accessing the devices controlled by Netbackup. Netbackup
requires exclusive
    access to its tape devices regardless of whether you use SSO or not.
 
Previous versions of Netbackup did not detect this problem and gave a false
impression that
backups succeeded when infact the data on tape had been overwritten because
during the
backup something else caused the tape drive to rewind. This would result in
an invalid backup and
corrupt the data on tape. Subsequent backups that attempted to use the same
tape would
result in continuous 85 errors which would normally point to a physical
media failure however this
was never the problem. The problem was that Netbackup expected the first
block on tape to be a
certain size and only allocated a buffer large enough to read it. However
the tape had been overwritten
(including the first block) with a larger block size (data block size). So
Netbackup would barf with
'no free space' because it could not read the first block into its small
buffer. This would result in 85 errors.
 
The moral of the story is never assume that I/O errors as reported by
Netbackup are due to failures with the
physical media.
 
In my experience, 9x40 tapes in open systems rarely fail due to media
imperfections. This is because open systems
timers are large enough to tolerate even the worst cases. 
 
The fact that this problem is detected in 34_3 and 4.5 is now a blessing.
 
Regards
Peter Marelas
 

-----Original Message-----
From: Lee Steichen [mailto:Lee_Steichen AT cnt DOT com]
Sent: Friday, 24 May 2002 12:57 AM
To: 'veritas-bu AT mailman.eng.auburn DOT edu'
Subject: [Veritas-bu] EDT-AdvanTAPE driver problems with AIX 4.3.3 &
NetBackup


NetBackup Datacenter v3.4 and v4.5 report status 84's on all backup jobs.
The writes appear to complete fine (correct number of files and KB's
written) though when NBU completes it reports status 84.  BPTM log reflects
that an "external event caused a media rewind". 
 
HBA - LP7000 (driver v4.1.0.16) (all IBM drivers have been removed)
 
AdvanTAPE driver version 2.2 on a AIX 4.3.3 RS6000 (F50)
 
9840 SCSI tape drives in a STK 9738 library
 
Pathlight FC/SCSI Gateway w/latest firmware
 
Veritas ovpass robotic control
 
AIX 4.3.3 is patched to maintanence level 4.3.3-4.3.3.0.10
 
Has anyone experience this situation? Are there any device driver settings
that need to be changed.  I am using the defaults.
 
 
Thanks, Lee
 
 

Lee Steichen 
Sr. San Integration Consultant 
Computer Network Technology 
763-268-6788 
lee_steichen AT cnt DOT com 

 


------_=_NextPart_001_01C202BB.01E240A0
Content-Type: text/html;
        charset="iso-8859-1"

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">


<META content="MSHTML 6.00.2715.400" name=GENERATOR></HEAD>
<BODY>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>This 
is a known problem with Netbackup irrespective of tape drive in 
use.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>See 
<A 
href="http://seer.support.veritas.com//docs/245505.htm";>http://seer.support.veritas.com//docs/245505.htm</A></FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>1. 
You 
want to ensure scsi reserve/release is enabled on all media 
servers.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>2. 
You 
want to ensure there is no device/tape monitoring software running on the media 
servers</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>&nbsp;&nbsp;&nbsp; that is accessing the devices controlled by 
Netbackup. 
Netbackup requires exclusive</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>&nbsp;&nbsp;&nbsp; access to its tape devices regardless of whether you 
use SSO or not.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002></SPAN><SPAN class=542035523-23052002><FONT 
face=Arial color=#0000ff size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>Previous versions of Netbackup did not detect this problem and gave a 
false impression that</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>backups succeeded when infact&nbsp;the data on tape had been overwritten 
because during the</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>backup 
something else caused the tape drive to rewind. This would result in an invalid 
backup and</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>corrupt the data&nbsp;on tape. Subsequent backups that attempted to use 
the same tape would</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>result 
in&nbsp;</FONT></SPAN><SPAN class=542035523-23052002><FONT face=Arial 
color=#0000ff size=2>continuous 85 errors which would normally&nbsp;point 
to&nbsp;a physical media failure however this</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>was 
never the problem. The problem was that Netbackup expected the first block on 
tape to be a</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>certain size and only allocated a&nbsp;buffer large enough to read it. 
However the tape had been overwritten</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>(including the first block) with a larger block size (data block size). 
So Netbackup would barf with</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>'no 
free space' because it could not </FONT></SPAN><SPAN 
class=542035523-23052002><FONT face=Arial color=#0000ff size=2>read 
the&nbsp;first block into its small buffer. This would result in 85 
errors.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>The 
moral of the story is&nbsp;never assume </FONT></SPAN><SPAN 
class=542035523-23052002><FONT face=Arial color=#0000ff size=2>that I/O errors 
as reported </FONT></SPAN><SPAN class=542035523-23052002><FONT face=Arial 
color=#0000ff size=2>by Netbackup are due to&nbsp;failures with 
the</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>physical&nbsp;media.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>In my 
experience,&nbsp;9x40 tapes </FONT></SPAN><SPAN class=542035523-23052002><FONT 
face=Arial color=#0000ff size=2>in open systems rarely fail due to&nbsp;media 
imperfections. This is&nbsp;because open systems</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>timers 
are&nbsp;large enough to&nbsp;</FONT></SPAN><SPAN 
class=542035523-23052002><FONT 
face=Arial color=#0000ff size=2>tolerate even the worst cases. 
</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>The 
fact that this&nbsp;problem is detected in </FONT></SPAN><SPAN 
class=542035523-23052002><FONT face=Arial color=#0000ff size=2>34_3 and 4.5 is 
now a blessing.</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2>Regards</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff size=2>Peter 
Marelas</FONT></SPAN></DIV>
<DIV><SPAN class=542035523-23052002><FONT face=Arial color=#0000ff 
size=2></FONT></SPAN>&nbsp;</DIV>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
  <DIV class=OutlookMessageHeader dir=ltr align=left><FONT face=Tahoma 
  size=2>-----Original Message-----<BR><B>From:</B> Lee Steichen 
  [mailto:Lee_Steichen AT cnt DOT com]<BR><B>Sent:</B> Friday, 24 May 2002 
12:57 
  AM<BR><B>To:</B> 'veritas-bu AT mailman.eng.auburn DOT 
edu'<BR><B>Subject:</B> 
  [Veritas-bu] EDT-AdvanTAPE driver problems with AIX 4.3.3 &amp; 
  NetBackup<BR><BR></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>NetBackup 
  Datacenter v3.4 and v4.5 report status 84's on all backup jobs.&nbsp; The 
  writes appear to complete fine (correct number of files and KB's written) 
  though when NBU completes it reports status 84.&nbsp; BPTM log reflects that 
  an "external event caused a media rewind".&nbsp;</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>HBA - LP7000 
  (driver v4.1.0.16) (all IBM drivers have been removed)</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT><FONT face=Arial size=2><SPAN 
  class=942144614-23052002>AdvanTAPE driver version 2.2 on a AIX 4.3.3 RS6000 
  (F50)</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>9840 SCSI tape 
  drives&nbsp;in a STK 9738 library</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>Pathlight FC/SCSI 
  Gateway w/latest firmware</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>Veritas ovpass 
  robotic control</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>AIX 4.3.3 is 
  patched to maintanence level 4.3.3-4.3.3.0.10</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>Has anyone 
  experience this situation?&nbsp;Are there any device driver settings that 
need 
  to be changed.&nbsp; I am using the defaults.</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN class=942144614-23052002>Thanks, 
  Lee</SPAN></FONT></DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <DIV><FONT face=Arial size=2><SPAN 
  class=942144614-23052002></SPAN></FONT>&nbsp;</DIV>
  <P><FONT size=4><STRONG><FONT face=Garamond>Lee Steichen</FONT> 
  <BR></STRONG></FONT><FONT face=Garamond>Sr. San Integration Consultant 
  <BR>Computer Network Technology <BR>763-268-6788 
  <BR>lee_steichen AT cnt DOT com</FONT> </P>
  <DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C202BB.01E240A0--