 Feedback
|
Table Of Contents
Maintaining and Managing the SN 5420
Identify the Location for Updated Software
Make Updated Software Available to the SN 5420
Set Updated Software as Boot Version
Backing Up System Configuration
Controlling SCSI Routing Services
Making Changes to Instance Configurations
Enabling and Disabling Connections
Viewing Operational Statistics
Maintaining and Managing the SN 5420
This chapter explains how to perform normal maintenance and management tasks associated with the Cisco SN 5420 Storage Router. The following tasks are covered:
•
Backing Up System Configuration
•
Prerequisite Tasks
Before performing any SN 5420 maintenance tasks, make sure you have configured system parameters as described in "First-Time Configuration" or "Configuring System Parameters".
Note
Installing Updated Software
The SN 5420 is designed to run on a continual basis without significant maintenance. However, from time to time, you may need to install updated software. The SN 5420 stores software images (along with configuration files, log files, and other information) on a local file system. This file system is stored on an internal, non-volatile Flash disk. The show software command displays a list of all software versions stored on the SN 5420, and the amount of disk space available for additional software.
Cisco.com provides registered users access to SN 5420 software updates. (See "Obtaining Technical Assistance" section for details on using Cisco.com.) Use a standard browser to download software updates and associated readme files to a location of your choosing. This download directory is the location from which you will make updated software available to the storage router.
Note
If you plan to use the CLI download software version command to make the updated software available to the storage router, the machine hosting the download directory must be running a web server. If the machine is not running a web server, use the SN 5420 web-based GUI to make the updated software available to the storage router. (See the online Help for details.)
The default location used for retrieval of updated SN 5420 software is set using the set software command and displayed using the show software command (Example 6-1). The show software command identifies the URL for the default download directory, the current version of software running on the storage router, and the version that will be used at system restart.
Example 6-1 Results of "show software" Command
[SN5420_A01]# show softwareVersion Boot Hash Sign Crash Size DateG2-0.9(34) N/A OK N/A 0 6.91 MB Jan 3 09:34 CST 2001G2-0.9(35) N/A OK N/A 0 6.94 MB Jan 12 14:30 CST 2001G2-1.0(00) N/A OK N/A 0 6.95 MB Jan 16 16:31 CST 2001Disk Space Free: 8.00 MbytesDownload URL: http://10.1.11.32/~software/sn5420Download User:Running Version: G2-0.9(35)Will boot Version: G2-1.0(00)To update SN 5420 software, perform the following steps:
Step 1
Step 2
Step 3
Identify the Location for Updated Software
You must specify the location from which to retrieve the updated software. If the default download location is not appropriate, you can reset it. Use the steps in the following table to identify the default download location.
If this location is not the one from which you would normally retrieve updated software, use the steps in the following table to set the appropriate download location.
Verify the new settings using the show software command, then save them using the save system or save all command.
Make Updated Software Available to the SN 5420
The download command makes a new version of SN 5420 software available to the storage router for boot purposes. Before attempting to install updated software, verify that sufficient space is available for the new software.
Use the steps in the following table to make a new version of software available to the SN 5420 Storage Router.
Timesaver
Use the steps in the following table to make updated software available to the storage router from a special location on a one-time basis.
Set Updated Software as Boot Version
Making new software available to the storage router does not change the currently running version of software, nor does it automatically set the new version to be booted at next system restart. You must take specific action to make the new software version bootable.
Setting software as the bootable version consists of verifying the software integrity and performing internal checks to ensure that the storage router can boot the specified version of software.
Use the steps in the following table to set the new software as the version to be booted.
When you set a new software version as the bootable version, internal checks are made to ensure that the new software can be run. In a cluster environment, the set software version command may temporarily suspend normal HA communications, causing a failover of any SCSI routing service instances active on this storage router.
Any instances with the primary attribute set to the IP address of this storage router will resume running on this storage router after it is rebooted. If you are not going to reboot the SN 5420 immediately, use the failover command to return the desired SCSI routing service instances to this storage router.
Note
Backing Up System Configuration
Backing up system configuration consists of saving selected SN 5420 configuration information to XML files that can be stored both locally and remotely. Should problems occur, SCSI routing services and access lists can be restored from these files. (See "Command Line Interface Command Reference," for details about what information is saved.)
While you can issue a save command at any time during a CLI command session, best practices suggest that you should back up your SN 5420 system configuration to file on a regular basis.
Configuration files are normally maintained in the savedconfig directory on the SN 5420. They can also be saved to a URL, allowing you to integrate the SN 5420 backups with other software archives.
Local Backup
Local backups allow you to store the resulting XML configuration file in the savedconfig directory on the storage router. Up to ten named configuration files can be locally stored.
Use the steps in the following table to perform a local backup that saves the configuration of all the current SCSI routing services to a file named mySCSI in the savedconfig directory.
Step 1
enable
Enter Administrator mode.
Step 2
save scsirouter all to mySCSI
Save all defined SCSI routing services to a file named mySCSI.
Remote Backup
Remote backups allow you to store the resulting XML configuration file to a user-defined location.
Use the steps in the following table to perform a remote backup that saves all configuration information to a file called config_11072000.xml, located at http://10.1.1.144/~backup/SN5420/Sys1.
Note
Restoring from Backups
SCSI routing services and access list configuration data can be restored from previously saved configuration files. You may choose to restore selected data, such as a specific instance of SCSI routing services, or all instances and access lists, using the restore from command.
The file from which configuration is restored must reside in the savedconfig directory. If you need to restore configuration data from a backup file existing elsewhere in the network, use the import CLI command to make the desired file available in the savedconfig directory.
Restoring configuration data copies all or part of the contents of the specified file into persistent memory; it does not change the running configuration for the storage router. Restored configuration data is only visible via the CLI show commands using the from bootconfig option, until the instance has been restarted or the system has been rebooted.
Restoring a Deleted SCSI Routing Services Instance
For example, suppose the SCSI routing services instance scsi1 was inadvertently deleted. Follow these steps to restore scsi1 from a configuration file that was saved to a URL.
Restoring an Existing SCSI Routing Services Instance
If you need to restore the configuration of a SCSI routing services instance that is still active in the storage router, you must stop the instance, restore the configuration from the selected file, then restart the instance. For example, follow these steps to restore the SCSI routing services instance scsi2 from the file scsi2_backup.
Restoring an Access List
When you restore an access list, existing entries are never deleted. The restore will add missing entries and overwrite entries of the same name, but will never purge or delete existing entries. If necessary, you can delete an entire access list and then restore if from a saved configuration file.
Follow these steps to restore the access list mylist1 from the file accesslist_backup.xml. In this example, mylist1 contains the following entries:
•
•
•
The saved access list in the configuration file accesslist_backup.xml contains these entries:
•
•
•
Resetting the System
There may be times when you need to return the SN 5420 configuration to factory defaults; for example, when moving a system between environments (like test and production), or for troubleshooting purposes.
To reset the SN 5420 Storage Router, perform the following steps:
Step 1
Step 2
Note
Step 3
Step 4
Reset All to Factory Defaults
In the example below, an existing SN 5420 is to be physically moved to another environment. It is not necessary to retain any current configuration information, because the system set up will be completely different.
Step 1
enable
Enter Administrator mode.
Step 2
clearconf
Clear the current system configuration, including network management information.
At the prompt, enter the Administrator password. Enter all to erase system configuration and management port settings, and all saved configurations and SCSI routing services (Example 6-2).
The SN 5420 will reboot.
Example 6-2 Reset SN 5420 Configuration
Enter admin password: *************************This process can restore factory default settings for the SN5420.Select "apps" to remove applications and retain system configuration settings.Select "system" to remove applications and system configuration settings.Select "named" to remove all named configurations.Select "all" to remove applications, system configuration and named configurations.The system configuration includes the management port, dns, admin and monitor login, ntp, and snmp. You will need to use the RS-232 terminal to reconfigure the management port if you erase the system configuration.The system will reboot if you select "apps", "system", or "all".Erase what? [apps/system/named/all/cancel (cancel)]
Note
Reset and Retain System Settings
In the next example, an existing SN 5420 is going to be used for testing purposes, and then restored to its current configuration. The storage router's system management information is not going to change for the test, so it may retained over the system reset.
Reset to Remove Saved Configuration Files
In the final example, a stand-alone storage router has joined a cluster and adopted the new cluster's configuration. Previously saved configuration files from the stand-alone period should be removed. The storage router's system configuration, management information, and SCSI routing services should remain unchanged.
Note
Controlling SCSI Routing Services
It is important to know where SCSI routing services are running. While automatic failover capabilities keep the SN 5420 cluster operational in times of system difficulties, manual HA controls provide the ability to distribute SCSI routing instances between the storage routers in the cluster to meet your specific network requirements.
The following are typical activities involved with controlling SCSI routing services in a cluster environment. While most of these activities are performed infrequently, some (such as viewing operational statistics) may be performed on a regular basis.
•
•
•
•
Making Changes to Instance Configurations
From time to time, you will make changes to the SCSI routing service instance configurations. Changes include adding or deleting a target, adding or deleting a LUN, remapping a target, or modifying access. It is important to understand the ramifications of these changes on the servers accessing the associated storage devices. For example, changing the instance configuration may change the device presentation to the server's iSCSI driver, effectively changing the name or number assigned to the device by the server operating system. Certain instance configuration changes, such as adding or deleting targets, adding or deleting LUNs within a particular target, or adding or deleting entire instances may change the order of the devices presented to the server. Even if the server is only associated with one SCSI routing service instance, the device order could make a difference.
Typically, the server operating system assigns drive identifications in the order they are received based on certain criteria. For example, a Linux system assigns drive identifications in the order they are received based on host, bus, target and LUN information. Changing the order of the storage device discovery may result in a changed drive identification. Applications running on the server may require modification to appropriately access the current drives.
If an entire SCSI routing service instance is removed, or there are no targets available for the server, the server's iSCSI driver configuration file must be updated to remove the appropriate reference before restarting the iSCSI driver. If a server's iSCSI configuration file contains a reference to an instance which does not exist or has no targets available for the server, the iSCSI driver will not complete a login and will not discover targets associated with any SCSI routing service instance.
For additional information and recommended procedures for changing iSCSI driver configuration, see the iSCSI driver readme and example configuration files.You can access the latest iSCSI drivers, readme and example configuration files from Cisco.com.
Enabling and Disabling Connections
A SCSI routing services instance becomes active, by default, once it is associated with a Fibre Channel interface to storage devices. Each target that is added to an instance is also, by default, enabled. However, no servers or hosts can connect or log in to that target because the target has no access list association. Once you associate an access list with a target, it is automatically enabled; the servers and hosts specified by access list entries are allowed to connect or log in to the target.
Use the set scsirouter target disable command to control access to the target without changing the access list association, or stopping the entire SCSI routing services instance. Existing connections and logins are not affected, but future connections and logins are prohibited.
Use the set scsirouter target enable command when you are ready to allow connections and logins again.
For example, suppose you have a problem with an entry in the access list webserver2. This access list is associated with the target webstorage2 which is, in turn, associated with the SCSI routing services instance foo.
Step 1
enable
Enter Administrator mode.
Step 2
show scsirouter foo stats
Display status, to confirm the SCSI routing services instance foo is active on this storage router.
Step 3
show scsirouter foo
Verify the name and current status of the target and access list. The target webstorage2 should be associated with the webserver2 access list and the target should be enabled. (Example 6-3).
Step 4
set scsirouter foo
target webstorage2 disableDisable access to the target webstorage2.
Example 6-3 Verify Target, Access List, and Target Status
[SN5420_PR1]# show scsirouter foo
foo primary "10.1.10.244"foo serverif ge2 10.1.10.234/32foo deviceif fc1foo target webstorage2 enabled "TRUE"foo target webstorage2 accesslist "webserver2"foo target webstorage2 lun 0 loopid "125" lun "32"Stopping & Starting Instances
If a storage router is experiencing a problem with a specific set of servers or storage devices, you may wish to stop the associated SCSI routing services instance from running anywhere in the cluster. The stop scsirouter command causes the specified SCSI routing services instance to cease running on the storage router, but does not cause a failover to another storage router in the cluster. This command effectively stops an instance from running anywhere in the cluster.
Once a SCSI routing services instance has been stopped, it can be re-activated by issuing the start scsirouter command. The start command must be issued from the same storage router as the stop command.
See "Command Line Interface Command Reference" for command details.
Viewing Operational Statistics
Use the show scsirouter stats command to display the status of the SCSI routing services instance, and to see the number of active connections and the number of logins that have occurred since the storage router was last restarted (or since statistics were last cleared).
For example, the show scsirouter command (Example 6-4) shows that SCSI routing services instance foo is currently active on storage router SN5420_PR1.
Example 6-4 Results of "show scsirouter" Command
[SN5420_PR1]# show scsirouter foo statsrouter status started version logins activefoo ACTIVE Jan 11 23:06:08 1.0.1 10 7Handling Failover
In a cluster, storage routers continually exchange HA information as heartbeats to detect failures in the cluster. To make sure that HA information is exchanged reliably between storage routers, the storage routers balance the transmission of heartbeats between the management and the high availability interfaces at half-second intervals. Failover of SCSI routing services is automatic when a storage router detects that the other SN 5420 in the cluster is no longer responding to heartbeats.
Each cluster supports up to four active SCSI routing service instances. Since each SN 5420 can also support up to four instances of SCSI routing services, high availability is ensured for each instance in the cluster (regardless of the division of those instances between storage routers).
Manual Failover
While failover of SCSI routing services is automatic, there may be times when you wish to manually move a SCSI routing services instance from one storage router to another. The move may be temporary, after which the instance will be moved back to its original location. At other times, you may want to make the move permanent, ensuring that the instance will continue running on the specified storage router whenever possible.
In the following examples, the cluster cluster1 is composed of two storage routers, SN5420Sys1 and SN5420Sys2. SN5420Sys1 is currently running instances scsi1 and scsi2, and is the primary storage router for both. SN5420Sys2 is currently running instances scsi3 and scsi4. The primary attribute for scsi3 and scsi4 is set to none, indicating no preferred storage router for either instance. (This is the default configuration for SCSI routing services.)
Failover as Temporary Move
Use the steps in the following table to move SCSI routing services scsi1 from SN5420Sys1 to SN5420Sys2, on a temporary basis.
Once the failover is complete, establish a Telnet session to SN5420Sys2 and verify that the SCSI routing services instance scsi1 is now running on that storage router, using CLI commands shown in Steps 1 and 2 of the previous table.
Note
Failover as Permanent Move
Use the steps in the following table to move the SCSI routing services scsi2 to SN5420Sys2 on a permanent basis.
Once the failover is complete, establish a Telnet session to SN5420Sys2 and verify that the SCSI routing services instance scsi2 is now running there, and that SN5420Sys2 is designated as the primary for that instance, via the show scsirouter scsi2 command.
Failover for Distribution Purposes
Due to a significant increase in traffic for scsi4, you decide to distribute all of the other instances (scsi1, scsi2, and scsi3) to the SN5420Sys1 storage router. SN5420Sys1 is already running scsi1 and scsi2.
Use the steps in the following table to move the SCSI routing services instance named scsi3 to SN5420Sys1.
Once the failover is complete, establish a Telnet session to SN5420Sys1 and verify that instances scsi1, scsi2, and scsi3 are now running there.
Note
