| PCI Hot Plug User's Guide I/O device edition - for Solaris(TM) Operating System - |
|
Contents
|
| Chapter 4 Network Devices | > 4.2 Extension of PCI cards |
This subsection describes the procedure extending PCI cards to the existing redundant configuration system with the following redundant configuration software.
The replacement and extension of PCI cards are supported only on Redundant Line Control Function of SynfinityLink.
Notes:
Because of the dynamic extension, if the environment of SynfinityLink or PRIMECLUSTER GLS has been modified, system reboot is not required.
When using other redundant software, please refer to manuals of each product.
Moreover, when extending PCI cards to the existing redundant configuration system, it is unnecessary to operate about the high layer products. Consequently, PCI cards can be extended without stopping the high layer products.
The hardware monitoring of the Machine Administration is stopped by using the following command.
| # /usr/sbin/FJSVmadm/prephp <Return> |
| # prtpicl -v > /tmp/prtpicl.pre <Return> |
Subsequent examples describe extension of PCI cards to the slot "R0B01-PCI#slot02." The "R0B01-PCI#slot02" indicates the I/O cabinet number : 0, the PCI / Disk Box number : 1, physical slot number : 02.
In the extension of a new redundant system, although two or more PCI cards are extended, whenever you add a PCI card, please perform from procedure "2.b" to "3. Connect PCI cards."
The relationship of the position of a PCI card and format of "Ap_Id" is shown below. (N: an integral number)
| Location of PCI slot | Element | Format of "Ap_Id" |
|---|---|---|
| Cabinet of PRIMEPOWER 900/1500/2500 |
Cabinet Number : X System Board Number : Y Physical Slot Number : ZZ |
|
| PCI / Disk Box | I/O Cabinet Number : X PCI / Disk Box Number : Y Physical Slot Number : ZZ |
|
| other than those above | Physical Slot Number : ZZ |
|
Please confirm that the PCI slot status which extended PCI cards is "empty unconfigured" by using cfgadm (1M) command.
| # cfgadm | grep R0B01-PCI#slot02 <Return> pcicsy18:R0B01-PCI#slot02 unknown empty unconfigured unknown |
| # cfgadm -x led=fault,mode=blink pcipsy18:R0B01-PCI#slot02 <Return> |
| # cfgadm pcipsy18:R0B01-PCI#slot02 <Return> Ap_Id Type Receptacle Occupant Condition pcicsy18:R0B01-PCI#slot02 unknown disconnected unconfigured unknown |
The cfgadm command is executed with the "configure" option to "Ap_Id," or the push button of the PCI bus slot is pushed. Then a extended PCI card is connected.
In addition, a push button is valid only in multiple user mode.
Please confirm that the slot status is "connected configured" by using the cfgadm (1M) command after the above-mentioned procedure.
| # cfgadm -c configure pcipsy18:R0B01-PCI#slot02 <Return> # cfgadm pcipsy18:R0B01-PCI#slot02 <Return> Ap_Id Type Receptacle Occupant Condition pcipsy18:R0B01-PCI#slot02 pci-pci/hp connected configured ok |
Please execute the following command in order to update configuration information, and restart the hardware monitoring of the Machine Administration.
| # /usr/sbin/FJSVmadm/postphp <Return> |
The extended PCI card is connected with network equipment by the cable.
The driver configuration is added by following operations.
Please save the result of the prtcicl (1M) command, and obtain the difference information between current result and the result taken at "2.a." Then, the driver instance number for the extended PCI card is obtained.
In the following example, since the instance number is 1, it can be determined that the interface name of the extended PCI card is "hme1."
| # prtpicl -v > /tmp/prtpic.post <Return> # diff /tmp/prtpicl.pre /tmp/prtpicl.post | more <Return> : > :status okay > :devfs-path /pci@8d,2000/network@1 > :driver-name hme > :binding-name SUNW,hme > :bus-addr 1 > :instance 1 > :_class obp-device > :name network : |
Please carry out this procedure, when it cannot identify relationship between an interface name and PCI bus slot by expanding two or more PCI cards especially.
| # /usr/sbin/FJSVmadm/inst2comp hme1 <Return> pcipsy18:R0B01-PCI#slot02 |
Please refer to each driver manual for details.
Please execute following commands in order to connect the extended PCI card to the existing redundant configuration system.
Notes:
On the NIC switching mode in Redundant Line Control Function, it is impossible to extend a new interface to the existing redundant configuration system. Moreover, when extending an interface on the GS/SURE linkage mode, it is necessary to extend an interface after deactivating the virtual interface (e.g. sha0) used by the GS/SURE linkage mode.
| Fast switching mode in Redundant Line Control Function # /usr/sbin/ifconfig hme2 plumb <Return> # /usr/sbin/ifconfig hme2 IP_address netmask + broadcast + -trailers up <Return> # /opt/FJSVhanet/usr/sbin/hanetnic add -n sha0 -i hme2 -f <Return> GS/SURE linkage mode in Redundant Line Control Function # /opt/FJSVhanet/usr/sbin/stphanet -n sha0 <Return> # /opt/FJSVhanet/usr/sbin/hanetconfig create -n sha2 -m n -i 192.168.20.10 -t hme2 <Return> # /opt/FJSVhanet/usr/sbin/hanetnic modify -n sha0 -t sha2,sha3,sha4 <Return> # /opt/FJSVhanet/usr/sbin/strhanet -n sha0 <Return> Multipath Function # /usr/opt/FJSViomp/bin/iompadm -c mpnet add mpnet0 hme2 <Return> # /usr/opt/FJSViomp/bin/iompadm -c mpnet start mpnet0 hme2 <Return> |
Notes:
It is necessary to configure the IP address specified by ifconfig(1M) command in /etc/inet/hosts and /etc/hostname.hme1. Unless these are configured, the virtual interface for fast switching mode can not be activated after rebooting the system. In addition, in order to use IPv6 address, please refer to the PRIMECLUSTER GLS (4.1A20 or later) manual.
Please switch an online communication path for extended communication path if needed. Please refer to manuals of each redundant configuration software about how to switch a communication path.
|
Contents
|