SSA Cabling and Adapter problem

From: Green, Simon (Simon.Green_at_EU.ALTRIA.COM)
Date: 11/17/04

  • Next message: Andrew Townsend: "Easy one! APARS -vs- Filesets" 
    Date:         Wed, 17 Nov 2004 15:49:59 +0100
To: aix-l@Princeton.EDU

    I've got an odd sort of cabling problem.

    Following some work to remove an HACMP cluster, one of our servers has been
    left with broken SSA loops.

    The current configuration has one Advanced Serial RAID Plus adapter
    connected to one 7133-D40:
    A1 -> Port 9
    A2 -> Port 16
    B1 -> Port 5
    B2 -> Port 4
    The drawer is set to BYPASS throughout. So the A loop is fine, but the B
    loop is broken, because the cables should have been connected to ports 1 and
    8.
    (There are actually six drawers, each set up as an independent loop; all are
    in the same condition.)
    The adapters have fast-write cache.
    The server is an S7A with two I/O drawers, running AIX 4.3.3.

    I can't see any way of correcting this without making some disks briefly
    unavailable. (Running cables between ports 1 & 8 would do it, but I don't
    have any short cables.)

    There is a seventh adapter in the server, not connected to any disks at all,
    so I thought I could use this as a sort of intermediary.
    I hooked up it's B ports to ports 1 and 8 on one of the drawers.
    When I checked the SSA Link Verification service aid for the original
    adapter, (ssa4), it showed a complete B loop, correctly identifying ssa6 in
    the middle of it. However, when I checked ssa6 it did not appear to have
    any disks connected to it, which worried me somewhat.
    Subsequently, when I tried removing the connections from ssa4, the disks
    went missing. Everything is mirrored across drawers: I wouldn't have
    attempted it otherwise. I put things back the way they were and got
    everything recovered swiftly.

    So: what should I have done to get ssa6 properly configured so that I could
    remove ssa4? (And repeat the procedure for the other five adapters.)
    Can anyone suggest an alternative procedure which does not involve making
    disks temporarily unavailable? 

    --
Simon Green
Altria ITSC Europe s.a.r.l.
AIX-L Archive at https://new-lists.princeton.edu/listserv/aix-l.html
<https://new-lists.princeton.edu/listserv/aix-l.html>
New to AIX? http://publib-b.boulder.ibm.com/redbooks.nsf/portals/UNIX
<http://publib-b.boulder.ibm.com/redbooks.nsf/portals/UNIX>
N.B. Unsolicited email from vendors will not be appreciated.
Please post all follow-ups to the list.
  • Next message: Andrew Townsend: "Easy one! APARS -vs- Filesets"

    Relevant Pages

    • Re: SSA Loop Verification Question (graphics)
      ... Indeed, all 5 nodes would need access to all disks at one time or another, ... many nodes in the loop. ... Print some cable plans off, and trace the loops with various combinations ... drawers as well. ...
      (AIX-L)
    • Re: SSA Loop Verification Question (graphics)
      ... many nodes in the loop. ... Be aware that if you have two nodes down there is a risk of some disks being ... Print some cable plans off, and trace the loops with various combinations of ... drawers as well. ...
      (AIX-L)
    • Re: Threading and Serial port issue
      ... I am using the Opennetcf Port class to handle the serial comms. ... I then sit in a loop waiting for data to arrive. ... private void SetupBoard_Click ...
      (microsoft.public.dotnet.framework.compactframework)
    • Re: Serial Communication Problem
      ... I have no problems reading from a serial port within a sub-thread. ... the basic loop within the thread, ... > The problem is that function WaitCommEvent never return! ...
      (microsoft.public.windowsxp.embedded)
    • Re: serial port question.
      ... Note that you have left the port in non-blocking mode. ... the termios settings are for raw mode and the VTIME and VMIN ... dependent upon that configuration, and when used as an example ... necessarily must be in a loop. ...
      (comp.os.linux.hardware)