"How to setup persistent name binding in Solaris? - Step by Step"

By WireINPUT -
Persistent Name Binding support is available for target devices. This support is provided for users to associate a specified device World Wide Port Name (WWPN) to a specified SCSI target ID.

For driver version earlier than 4.15, it is implemented through both the /kernel/drv/sd.conf and /kernel/drv/qla2300.conf files. The matching between the target device port, the SCSI target ID, and the adapter driver instance happens inside /kernel/drv/qla2300.conf.

For driver versions 4.15 and later, the driver dynamically discovers targets, so no entries in the sd.conf or st.conf files are required. Note that if you disable dynamic discovery of devices (via the hbaX-disable-dynamic-device-support parameter), you must populate the sd.conf and st.conf files as noted above for pre-4.15 drivers.

The Persistent Name Binding support for target devices associates a SCSI target ID to a specified device World Wide Port Name or port ID. If port ID binding is specified it takes precedent over World Wide Port Name binding. In other words, this entry changes the target ID of the specified device to the number specified in the entry.

To enable persistent binding by WWPN, lines of the following format should be added in the qla2300.conf file for each target device:

hba<#>-SCSI-target-id-<#>-fibre-channel-port-name="";

For example, to associate the following WWPN with target ID 2 on adapter instance 1, the entry is:

hba1-SCSI-target-id-2-fibre-channel-port-name="2100002037004ac9";

To enable persistent binding by port ID, lines of the following format should be added in the qla2300.conf file:

hba<#>-SCSI-target-id-<#>-port-id="";

For example, to associate the following port ID with target ID 2 on adapter instance 1, the entry is:

hba1-SCSI-target-id-2-port-id="10031e";

The current effective mapping between device WWPN or port ID and SCSI ID of the target devices detected by a particular adapter instance can be found in the /var/adm/messages file. The information looks similar to this:

qla2300: hba1-SCSI-target-id-2-fibre-channel-node-name="2000002037004ac9";
qla2300: hba1-SCSI-target-id-2-fibre-channel-port-name="2000002037004ac9";
qla2300: hba1-SCSI-target-id-2-port- id="10031e";
qla2300: hba1-SCSI-target-id-3-fibre-channel-node-name="2000002037004ac7";
qla2300: hba1-SCSI-target-id-3-fibre-channel-port-name="2000002037004ac7";
qla2300: hba1-SCSI-target-id-3-port-id="10031f";

IMPORTANT
Unlike other parameter entries in the qla2300.conf file, values indicated in persistent name binding entries for instance 0 do not apply to other instances when there are no entries for the other instances.

Persistent Name Binding entries are optional entries in the qla2300.conf file. These entries do not dictate which devices are scanned by the driver, and do not perform any device masking. Absence of this entry for any given device name or SCSI ID does not direct the driver to perform any type of device masking.

For target devices, if no Persistent Name Binding entries are specified, the default SCSI target IDs assigned are the loop ID of the devices. Please refer to the FAQ section for which default SCSI target IDs are assigned to fabric devices.

Comments

Post a Comment

Popular posts from this blog

VMware PSOD Purple Screen of Death - Debugger waiting (world 2078) -- no port for remote debugger. "Escape" for local debugger

By WireINPUT -
Image
How to deal with PSOD - VMware PSOD Purple Screen of Death - Debugger waiting (world 2078) -- no port for remote debugger. "Escape" for local debugger - Easy troubleshooting steps and how to identify or determine exact root cause of the PSOD issue Host is down with PSOD with the above said error. Verified BIOS settings, worked with on site engineer, contacted the IBM server support (since this is IBM server, have contacted IBM, if it is IBM or some other hardware, you may need to reach out the respective vendor) Sometimes it could be server hardware issue like DIMMs, CPU, Processor related or may be a chance of network related. Please have a through check on the above said and come to a conclusion what could the issue. Based on that, need to troubleshoot the issue.
Read more

The Windows Time Service terminated with the following error - Event ID 7023 & 46

By nag -
After you upgrade a Microsoft Windows Server 2003-based domain controller to Windows Server 2003 Service Pack 1 (SP1), the Windows Time service may not start. In this scenario, the following events may be logged in the Windows System log. Message 1 Event Type: Error Event Source: Service Control Manager Event Category: None Event ID: 7023 Description: The Windows Time service terminated with the following error: Not all privileges referenced are assigned to the caller. For more information, see Help and Support Center at http://support.microsoft.com. Message 2 Event Type: Error Event Source: W32Time Event Category: None Event ID: 46 Description: The time service encountered an error and was forced to shut down. The error was: 0x80070700: An attempt was made to logon, but the network logon service was not started. Additionally, when you try to start the Windows Time service manually, you may receive one of the following error messages: Error 1083: The executable program that this servic
Read more

IBM x3650 M4 Series Server Model - Activation Keys Backup to be taken for IMM Moduel II, why?

By WireINPUT -
Image
If you are planning to replace the system board (Mother Board) for M4 series of IBM xSeries 3650 model server, you should and must take the activation keys backup, otherwise you will miss the IMM key. IMM key something like iLO Activation key on HP servers. If that  iLO key didn't acivated, then you won't get the 'Remote Access' option. In the same way, IBM series requires a key to be activated to access the IMM console. IMM Activation to be backuped and restore once the new board get replaced. Here is the screenshot there you see more options what we discussed so far -
Read more