Netra T1 (200) and Sun Storedge D1000
Following scenario:
I want to set up an Sun Cluster 3 enviroment. For that I have two Netra T1 (200) and two D1000.
The Netra T1 has only one SCSI interface on borad, so I added an PCI SCSI/Ethernet combocard in both server. Now I am able to write data from each server to each disk array.
But the following error occur:
After all parts were connected, I execute an "probe-scsi-all" on one server to see the disks in the D1000, but I cannot see any disks of the D1000.
On the second server I got the error message:
Nov 5 10:58:13 wapgw2 scsi: WARNING: /pci@1f,0/pci@1/pci@5/SUNW,isptwo@4 (isp0):
Nov 5 10:58:13 wapgw2 SCSI Cable/Connection problem.
Nov 5 10:58:13 wapgw2 scsi:Hardware/Firmware error.
Nov 5 10:58:13 wapgw2 scsi: WARNING: /pci@1f,0/pci@1/pci@5/SUNW,isptwo@4 (isp0):
Nov 5 10:58:13 wapgw2 Fatal error, resetting interface
I work with Solaris 8, patchlevel 24. The Prom was already updated (111991-06).
Can somebody help me?
[1024 byte] By [
one_4_all] at [2007-11-25 23:13:02]

# 3
Hello,
thanks for help.
In order to change the scsi initiator on one of the two machines, I need the path of both interfaces.
Since now I changed it not.
As I know you do the following:
- switch off one server
- set auto-boot? to false
- reset-all
- probe-scsi-all
after that you have the path of both scsi interfaces. With that info you perform a nvedit and set the id of one machine for example to 6.
But I thought that regardless of the scsi id, I should be able to do a probe-scsi-all.
Was this a mistake from me?
# 4
Additional Info:
The netra t1 has only one pci slot.
For sun cluster 3 I need 3 ethernet ports, two for the cluster interconnects and one for the public network.
The netra has only one scsi interface one board. But I need two scsi interfaces, one to connect to D1000-A and one to D1000-B which are mirrored.
For that situation there exists a PCI combo card for the netra t1, which has one 10/100 MBit Ethernet Port and one SCSI Interface.
That was I mentioned with c"ombocard"in my explanation.