T.R | Title | User | Personal Name | Date | Lines |
---|
152.1 | LAVc is Cluster too | HYEND::BLYONS | VAXcluster SASE | Tue Oct 09 1990 10:36 | 10 |
| As mentioned in the announcement of FDDI, VAXclusters will (some day)
be able to use that media as a cluster interconnect. LAVc is rather
an obsolete term for (pre-V5 of VMS) VAXclusters using the NI as its
interconnect. Since V5, a cluster is a cluster no mater what its
interconnect.
So, to answer the question... yes and no. You will have Clusters
on the FDDI but they probably will not be called Local Area VAXclusters.
Bob L.
|
152.2 | Cluster Support
Cluster Support | SAC::GOODINGG | | Thu Jan 10 1991 12:36 | 65 |
|
I'm currently looking at the rewiring of a Digital building which will
start to take place in the next couple of months. We are looking at the
feasability of using an FDDI backbone. The building will contain a
large number of clusters, whose members will be distributed throught the
building. It is therefore essential that the networking option we
choose will support distributed clusters running old(V5.0-V5.3) and new
versions of VMS.
Configuration 1
All VAX'S are part of the same cluster. Therefore can clusters be
supported when the VAX'S are attached to ethernet but there is a FDDI
ring in between the ethernet segments. Is this configuration possible
and supported today?.
+---+ +---+ +----+
|VAX| |VAX| |VAX |
+-+-+ +-+-+ +-+--+
+---+---+ ETHERNET | ETHERNET
| |
+-----+-------+ +------+------+
|10/100 Bridge| |10/100 Bridge|
+-----+-------+ +-----+-------+
| |
|FDDI | FDDI
+--+-+ FDDI RING +--+-+
|CONC+------------------+CONC|
+--+-+ +-+--+
| |
+----------------------+
FDDI RING
Configuration 2
All VAX'S are in the same cluster, but the VAX's are directly joined by
FDDI. I assume that this configuration will only be possible when VMS has
be upgraded to support the FDDI adapters.
+---+ +---+
|VAX| |VAX|
+-+-+ +-+-+
| |
FDDI +------------+ +-----------+ FDDI
| |
+-+--++
|CONC |
+-----+
OR
+---+ +----+
|VAX| |VAX |
+-+-+ FDDI RING +--+-+
| +-----------------+ |
| | | |
| +--+-+ +--+-+ |
FDDI +---+CONC| |CONC+---+ FDDI
+--+-+ +--+-+
| |
+-----------------+
Thanks
Gary
|
152.3 | Pointer | AKO569::JOY | Get a life! | Thu Jan 10 1991 13:30 | 6 |
| I believe both configurations will be supported when VAXcluster support
over FDDI is announced, but that won't be until early FY92 most likely.
You might want to ask this in the VLAB::FDDI_CLUSTERS conference.
Debbie
|
152.4 | | KONING::KONING | NI1D @FN42eq | Thu Jan 10 1991 13:51 | 4 |
| Since the first configuration involves only VAXen connected to the Ethernet,
it should be supported today.
paul
|
152.5 | FDDI backbone and 10/100 bridges work now | HYEND::KPARRIS | SET HOST, then Control-P | Thu Jan 10 1991 14:58 | 4 |
| VAXes on Ethernet with an FDDI backbone is supported today, and in fact my
satellite workstation connects to the rest of the cluster over an FDDI backbone
in MR01. The SPD does limit the size and/or number of nodes which have only
Ethernet connections between them.
|
152.6 | Limit of nodes | SAC::GOODINGG | | Fri Jan 11 1991 07:32 | 6 |
| Can you point me at the SPD or give me more details on this size and/or number
limit of the nodes which can have only ethernet connections between them
Thanks
Gary
|
152.7 | VAXcluster SPD info. | CVG::TOMASWICK | Joe - VAXcluster Validation | Fri Jan 11 1991 09:36 | 15 |
|
The current VAXcluster SPD is V5.4 SPD # 29.78.04.
You can get a copy of it through $VTX SPD.
You can connect cluster nodes through FDDI bridges as long
as you follow the standard configuration rules. (ie. no more than 7
levels of bridges (FDDI+Ethernet combined)).
True FDDI based clusters (ones with direct VAX connections)
may have distance length and number of stations allowed on the FDDI
ring restrictions. There will be more infomation on this in the next
few months as configuration guidelines are developed for FDDI based
clusters.
Joe
|
152.8 | Node size/number limits from SPD | HYEND::KPARRIS | SET HOST, then Control-P | Fri Jan 11 1991 10:03 | 17 |
| Re .6:
The relevant portion of the SPD is:
o A CI path is needed between any pair of the following CPUs in
a VAXcluster system. An Ethernet only VAXcluster system may
include a maximum of one of the following CPUs:
- VAX 6000-xxx
- VAX 85xx
- VAX 86xx
- VAX 8700
- VAX 88xx
- VAX 9xx
Undoubtedly the introduction of FDDI will mean this section of the SPD must be
updated.
|
152.9 | Whither VLAB::FDDI_CLUSTERS? | ERICG::ERICG | Eric Goldstein | Tue Jul 21 1992 05:09 | 5 |
| .3> You might want to ask this in the VLAB::FDDI_CLUSTERS conference.
For some time now, I have gotten "remote node not reachable" when I try to open
this conference. Has the conference moved, or have it and its node both gone
away?
|
152.10 | FDDI_CLUSTERS was deleted | CVG::TOMASWICK | Joe - VAXcluster Validation/aka CVG | Tue Jul 21 1992 09:09 | 6 |
| -< Whither VLAB::FDDI_CLUSTERS? >-
Any FDDI VAXcluster discussions are on SPEZKO::CLUSTER. The
VLAB::FDDI_CLUSTERS file was considered redundant and was deleted.
Joe
|
152.11 | FDDI_CLUSTERS rises from the grave. | ERICG::ERICG | Eric Goldstein | Wed Jul 22 1992 02:39 | 5 |
| .10> Any FDDI VAXcluster discussions are on SPEZKO::CLUSTER. The
.10> VLAB::FDDI_CLUSTERS file was considered redundant and was deleted.
Thanks for the information. This morning, interestingly, VLAB::FDDI_CLUSTERS
is back on line. I guess that someone read .9 and decided to restore it.
|
152.12 | | STAR::PARRIS | _ 13,26,42,96... What comes next? | Fri Jul 24 1992 10:07 | 2 |
| VLAB:: is now only one node within the SCHOOL:: cluster alias. You may have
better luck accessing conferences which were on VLAB:: on SCHOOL:: instead.
|
152.13 | re .12: Thanks. | ERICG::ERICG | Eric Goldstein | Sun Jul 26 1992 02:47 | 0
|