I want to thank everyone who responded. We now have redundant connections for all three CTCA interfaces between the two systems, and shouldn't have to worry if this ever comes up again....
Also, Alan: The ISFC links do show as one being up and the other down. Do you know if it will automatically switch to the back-up connection should the first one fail, or will it take manual intervention? For anyone interested, here are queries of the three connection sets... q islink Link: E242 Type: CTCA Node: POLAR Bytes Sent: 87043 State: Up Bytes Received: 60631 Buffer Count: 16 Status: Idle Link: E342 Type: CTCA Node: Bytes Sent: 20732 State: Down Bytes Received: 20732 Buffer Count: 16 Status: Idle Ready; T=0.01/0.01 14:37:13 smsg rscs q group polar Ready; T=0.01/0.01 14:42:22 Group Alternate Name --------------- Primary Links -------------- Link POLAR CTC1 CTC2 ... ... ... ... smsg rscs q links Ready; T=0.01/0.01 14:42:30 Link Line Name Status Type Addr LU Name Logmode Queueing CTC1 connect NJE E240 ... ... priority CTC2 connect NJE E340 ... ... priority *NOTHERE inactive NOTIFY 0000 ... ... FIFO *UNKNOWN inactive NOTIFY 0000 ... ... FIFO *LPRHOLD inactive NOTIFY 0000 ... ... FIFO *UFTHOLD inactive NOTIFY 0000 ... ... FIFO 6 links found smsg pvm q system Ready; T=0.01/0.01 14:44:14 08/17/10 14:44:14 DVMQRY080I The local node is GRIZZLY, Users 0, PVM 2.1.1901, built 10/26/99 14:12:10 08/17/10 14:44:14 DVMQRY081I Link CTC1 Type CTCA Address 141 Users 1 CONNECT Group POLAR 08/17/10 14:44:14 DVMQRY081I Link CTC2 Type CTCA Address 241 Users 0 CONNECT Group POLAR -- Robert P. Nix Mayo Foundation .~. RO-OC-1-18 200 First Street SW /V\ 507-284-0844 Rochester, MN 55905 /( )\ ----- ^^-^^ "In theory, theory and practice are the same, but in practice, theory and practice are different." On 8/17/10 10:17 AM, "Alan Altmark" <alan_altm...@us.ibm.com> wrote: > On Tuesday, 08/17/2010 at 09:52 EDT, RPN01 <nix.rob...@mayo.edu> wrote: >> I?m going to dig into the manuals in a moment, but can anyone quickly > tell me >> if the CTCA connections used by the various parts of CSE can have > redundant >> sets of connections? (I?m about to lose the set I?m using to a POR, and > I need >> to know if I should switch to a second set, or do a better job of >> implementing...) >> >> RSCS uses a set, PVM uses a set, and we have one set for ISFC. Can any > or all >> of these have duplicate paths across a second set of CTCAs? > > ISFC does not allow a redundant connection. If memory serves, you can > configure them, but the duplicate will not be activated. You have to > deactivate one before another can go active. If you have alternate paths, > then you will still maintain the integrity of the cluster while you take > the links down/up. Just do them one at a time. > > RSCS and PVM allow redundant paths, but you have to ensure that you have > created groups and/or routing tables to handle them. > > Alan Altmark > z/VM Development > IBM Endicott
