Storage Automated Diagnostic Environment> for the first time, or to make changes, to ensure proper monitoring, notification, and diagnosis.
Once you have configured the Storage Automated Diagnostic Environment Master and Slave hosts, you should use the Maintenance section only to update email notifications or to update the monitored devices.
When the Storage Automated Diagnostic Environment> package is installed on a host, it registers with the Master agent and an entry is added to this page. For this automatic registration to work, you must first install and run the Master agent. A 'Y' will appear in the Active field of the Maintain Hosts window. Once all hosts are installed, you can access this page to change the configuration of each host.
The three main functions of the Maintain Hosts window are:
- Add a New Slave - Use when the agent does not automatically register. This can occur if a previously-installed instance is temporarily deleted from the configuration. Note that the Slave is not active if manually added.
- Ping Slaves - Verify that the agent at each host is up and running. The master will ping the slaves and request their hostid to verify that the host is up and that the Storage Automated Diagnostic Environment> service is available.
- Push Slave Configurations - Once all the slaves have been updated, you can push the new configurations to the respective master by using Push Slave Configs. Each instance of the Storage Automated Diagnostic Environment has identical copies of the configuration file, but the administration of the Slaves is done from the Master.
Maintain Site Information
| [ Up ] |
From the General Maintenance window, select Site Information. The fields on this page must be completed before you proceed. The fields with an asterisk (*) are mandatory. To update the master configuration, click an existing Host Name link from the Maintain Hosts window and update the fields as required, then click Submit.
If a slave agent does not automatically register, you can add a new slave manually.
To access the Add New Slave window:
- Click the Maintain Hosts link on the Storage Automated Diagnostic Environment> window.
- Click the Add a new Slave button on the Host List eng window.
The Maintain Hosts: Add New Slave window is displayed.
Note: The Categories to Monitor choices indicate Sun Microsystem network storage
devices that are supported with this instance of the Storage Automated Diagnostic Environment>.
- Type the 8-digit, alphanumeric HostID, which is the same as the system ID in the /usr/sbin/hostid directory. You can use the hostid command on the Host to obtain a valid HostID.
- Type the name of the Host (hostname.domain).
- Select the Category or Categories to Monitor. The categories you specify are supported with this instance of the Storage Automated Diagnostic Environment.
- From the Monitoring Frequency drop down menu, select a variable between five and 120 minutes. This frequency variable affects how frequently the cron executes. The default is 5 minutes.
- The Alternate Master checkbox is enabled the first time you add a new Slave. For the Slave to be an alternate Master, you must disable this checkbox.
- Enable the Sun StorEdge 3900 and 6900 series button if the device or devices the Storage Automated Diagnostic Environment will monitor are part of a Sun StorEdge 3900 or 6900 series solution.
- Enable the Monitor In-Band Data Path button if you want to view devices on the topology of a Host that are accessible in-band by that host.
- The default Message Log path is /var/adm/messages, which displays the history of the status messages.
- The T3 message log path is to where the Sun StorEdge T3 and T3+ array logs are sent. You can check and verify the path by looking at the /etc/syslog.conf on the Host.
- Click Add to add the new Slave Host.
Caution:If the system you are adding as a Slave is currently a Master, the following message is displayed: WARNING: System is currently configured as a Master. Please use the Push Slave Configs section to change this system to a Slave.
- To automatically update the Slave's configurations and to verify that all Slaves agree with the Master configurations, click Push Slave Configs from the Maintenance menu.
Notes:
- Enter an 8-digit HostID in the /usr/sbin/hostid directory. The Storage Automated Diagnostic Environment> will not add a new slave without it.
If you do not know the HostID, you can enter a dummy number and later correct it. Note, however, that the HostID cannot match any other previously-accepted HostID.
The Host you add as a Slave must be accessible over the Ethernet from the Master and must, at a minimum, be able to respond to a ping.
- If you manually enter a Slave Host, and no instance of the Storage Automated Diagnostic Environment> is installed on that host, you must also manually enter the HostID for that Slave host.
Otherwise, the Categories to Monitor field will not have any devices selected, and the Storage Automated Diagnostic Environment> will not be active.
The Active state will depend on the Slave configuration file (rasagent.conf.push) in the /var/opt/SUNWstade/DATA directory, and will also depend on whether the Active flag is set. If there is simply no Slave configuration file (for example, if the Host is a Master), the Active state will be set to OFF.
You can request the Storage Automated Diagnostic Environment> to probe the environments for the device types shown on the Device Discovery window.
- The Storage Automated Diagnostic Environment> discovers the Sun StorEdge A5000 arrays, the Sun StorEdge A3500-FC Raid Controllers, the Sun Fire V880 Disk, the Sun StorEdge D2 array, Internal Fibre Channel Disk, and Fibre Channel Tape devices on the physical data path attachments to the Host.
- The Storage Automated Diagnostic Environment> automatically discovers the Sun StorEdge network FC switch-8 and switch-16 switches by reading the /etc/fcswitch.conf file (the format is similar to etc/hosts). Additionally, the Subnet functionality asks for an ip network prefix to scan for Sun StorEdge Network 2 GB Switch-16 switches and Brocade Silkworm 1 GB and 2 GB switches answering to SNMP queries.
- The Storage Automated Diagnostic Environment> automatically discovers the Sun StorEdge T3 and T3+ arrays by scanning the Sun StorEdge T3 and T3+ array logfile on all Hosts. Entries that contain a valid IP name or address are added to the configuration.
- The Storage Automated Diagnostic Environment> discovers the Sun StorEdge 3900 and 6900 series using the /etc/host file of the Storage Service Processor.
Note:
As long as there are valid Sun StorEdge T3 and T3+ array entries in this log file, the auto-discovery will find them and add them.
You can avoid adding older, unwanted Sun StorEdge T3 and T3+ arrays by rolling over the log file, much the same way the system log rolls over.
Reference the scripts /usr/lib/newsyslog and /etc/cron.d/logchecker.
In a cascaded switch-configured SAN, only the switch with the Host attachment (that is, the local switch) needs to be in the /etc/fcswitch.conf file. The Storage Automated Diagnostic Environment> will discover the cascaded switch when the first switch is discovered.
Note:
You must manually add Brocade Silkworm switches, or search the LAN for a Brocade Silkworm switch and then manually delete the unwanted devices which were discovered elsewhere on the specified LAN.
Discover a Sun StorEdge 3900 or 6900 Series Cabinet
| [ Up ] |
Sun StorEdge 3900 series storage systems for direct attach storage (DAS) environments, and Sun StorEdge 6900 series storage systems for Storage Area Network (SAN) environments, are complete, preconfigured storage solutions. End users have only restricted accessibility to the internal components of the subsystems, but they do have full utilization of its storage functionality.
Select the Sun 3900/6900 button on the Device Discovery window to instruct the Storage Automated Diagnostic Environment to query for a Sun StorEdge 3900 or 6900 series cabinet. The IP address or name of the service processor in the Sun StorEdge 3900 or 6900 series cabinet should be selected. The StorADE master will then query locally or the slave address selected for the 3600/6900 subsystem.
- Select the host from the Run on Host menu.
- Enable the Turn Monitoring On toggle button.
- Click Confirm 3900/6900 Discovery.
Note:
The discovery of Sun StorEdge 3900 and 6900 Series subsystems is dependent on the existence of the Solution.golden configuration file which resides in the SUNWstade /var/opt/SUNWstade/DATA directory. This file is created by running the Storage Automated Diagnostic Environment config_solution(1M) utility which is located in the SUNWstade /bin directory. This utility should always be run during manufacture, upgrades and component replacements.
Note:
The Storage Automated Diagnostic Environment> employs a direct Ethernet connection (out-of-band) to the Sun StorEdge T3 and T3+ arrays, virtualization engines, Sun StorEdge network FC switch-8 and switch-16 switch devices, and Brocade Silkworm switches.
Search the LAN for Sun StorEdge Devices using Subnet
| [ Up ] |
Use the Subnet button to search the LAN for Sun StorEdge network FC switch-8 and switch-16 switch devices, Brocade Silkworm switches, and Sun StorEdge T3 and T3+ arrays.
- Enter the IP address of the subnet you want searched.
- Enter the Subnet prefix for the search.
The Subnet functionality will search the subnet for valid devices.
The search results are sent back to the Master agent. The Master agent moves the recently discovered, valid devices that were previously in the subnet into the main configuration file.
Set Up Sun StorEdge T3 and T3+ Array Message Monitoring
| [ Up ] |
In order for the Storage Automatic Diagnostic Environment to monitor messages from a Sun StorEdge T3 or T3+ array, you must mirror the
Sun StorEdge T3 or T3+ array's /syslog to a Host with the Storage Automated Diagnostic Environment> installed and
configured to monitor Sun StorEdge T3 and T3+ arrays.
- Refer to the Sun StorEdge T3 and T3+ Disk Tray Installation, Operation, and Service Manual for
procedures on how to set up the Sun StorEdge T3 array and the host to forward syslog messages.
- Refer to Maintain Hosts section to configure each host with the name given to the T3 and T3+ array message log file.
To add a new device:
- Click the Add New Device link on the General Maintenance Menu.
- Complete the fields using the descriptions that follow.
- Device Name
Type the device name the Storage Automated Diagnostic Environment> will use to identify the device.
- Telnet Password (optional)
Type the telnet password. This field is optional for the Sun StorEdge T3 and T3+ array.
For a more detailed analysis, enter the root password for the Sun StorEdge T3
array, which enables a higher level of accessibility to the Sun StorEdge T3 array.
- Primary Host
Select the Storage Automated Diagnostic Environment's primary host that will perform the actual monitoring
of the device.
Note:
The default is the Host that automatically discovered the device.
- IP Name/Address
Type the IP address that identifies the host to the network.
Once your list is complete, click Update Selected Slaves from the Push Configs link.
The Update Selected Slaves functionality is required to update a Slave's configurations after you have added or updated one or more devices.
Note:
To find the device name for the Sun StorEdge A3500FC array,
use the RM6 command drivutil -l devName, where devName is derived
from the RM6 command /usr/lib/osa/bin/lad.
Use the Maintain Devices window to configure the Host to monitor each agent device. You can also use this window to change the name that the Storage Automated Diagnostic Environment> will use as a reference to that device through its email notifications.
Be aware of the host's access method when making decisions about
which devices the Storage Automated Diagnostic Environment> will monitor.
- Monitor Sun StorEdge A3500-FC arrays from a Host that has an attachment through the Fibre Channel cables and can run the Sun StorEdge
RAID Manager commands.
- Monitor Sun StorEdge A5000 arrays, Sun Fire V880 Disk, and Sun StorEdge D2 arrays from the Host that has a direct
attachment through the fibre channel cables.
- Monitor Sun StorEdge network FC Switch-8 and Switch-16 switches on any host on the subnet, because the probing is done over the network.
- Monitor Brocade Silkworm switches on any host on the subnet, because the probing is done over the network.
- Monitor Sun StorEdge T3 and T3+ arrays from the host that has access to the t3 message log file and has an Ethernet connection on the subnet.
.
- Monitor Internal Fibre Channel Disk and Fibre Channel Tape devices from the Host that has an attachment through the Fibre Channel cables.
- Monitor Sun StorEdge 3900 and 6900 series cabinets from the Storage Service Processor, which has Ethernet connection to the Sun StorEdge T3 and T3+ array, the Sun StorEdge network FC switch-8 and switch-16 switches, and the virtualization engine.
Note:
The Primary Host column lists the registered host that will do
the actual device probing.
The buttons indicate Sun StorEdge devices that are supported with this instance of the Storage Automated Diagnostic Environment>.
Manually Update a Device
| [ Up ] |
- Click the Maintenance link on the Storage Automated Diagnostic Environment main window.
- Click the Maintain Devices link on the General Maintenance menu.
- From the Maintain Devices window, select the device you want to update from the Device Name column.
- Once you have corrected the necessary fields, click Update.
- After you have added a device or updated a device, refresh the system by clicking Re-Discover.
Notes:
- Changing the name of the device changes only the reference to that device
within the Storage Automated Diagnostic Environment>. It does not affect the access or reference to
that device within the operating system.
- If you change the configuration of a device (for example, if you change the
chassis ID of a switch), you must delete that unwanted device. Storage Automated Diagnostic Environment will display a message
that the previous device has been removed. Topology will not, however,
display the deleted device until you manually add and discover the device.
You can delete an unwanted device once the device has been removed from the site or if device monitoring is no longer needed.
To delete a device:
- Click the Maintenance link on the Storage Automated Diagnostic Environment main window.
- Click the Maintain Devices link on the Maintenance menu.
The Maintain Devices window is displayed.
- From the Maintain Devices window, select the device you want to delete from the Device Name column.
The Maintain Host: Update Device window is displayed.
- Click Delete.
Deleting a device from the configuration will not remove the access to the device instrumentation in cache immediately. This will, however, be cleared the next time the agent runs. State information for that device will also be maintained until the agent's next run. Therefore a device can be removed from a configuration and still be viewed in the Topology and Instrumentation as a snapshot in time, until you once again execute the Storage Automated Diagnostic Environment>.
If you rename a device and then execute the agent from the command line or from the GUI, the Storage Automated Diagnostic Environment displays a message that the previous device has been removed. The Storage
Automated Diagnostic Environment Topology will not, however, display the renamed device until you rediscover the device.
You can avoid this problem by manually adding the world wide name (WWN) into the Storage Automated Diagnostic Environment
configuration file, which is located in the file /var/opt/SUNWstade/DATA/rasagent.conf.
Note:
Unless you remove the SUNWstade directory or perform a clean ras_install, the configuration file remains on the system between upgrades.
Customizing Email Deliveries
| [ Up ] |
You can use the Email Notification window to customize the generation of emails to yourself or to other administrators at their companies. For example,
if you are interested in receiving only high-priority alerts coming from Sun StorEdge T3 and T3+ arrays, you can create a specialized notification for this instance only.
Alerts are sent only to valid email addresses that you have entered through the Email Notification function.
Local notification does not send mail to the provider.
You can customize the following local notification information:
- Notification type
- Email address
- Category
- Event Type
- Priority
Note:
The local email/pager notifications feature is optional and does not affect the main transmission functions of the Storage Automated Diagnostic Environment>.
The Master instance of the Storage Automated Diagnostic Environment> is the only instance generating emails based on local notifications.
Slave instances of the Storage Automated Diagnostic Environment> send their alerts to the Master, which filters them and forwards them to the providers, if selected, and to local system administrators, if configured.
Local Notification: Add New Entry
| [ Up ] |
Create a notification record for each system administrator who would like to be notified of the Storage Automated Diagnostic Environment> findings.
You can restrict the findings by Event Type, Priority, and Device Category. Multiple events can be selected and sent to an email or a pager. This flexibility enables a customized alert notification scheme.
To add a new entry:
- Click an email link to update or delete an existing email address.
- Click the Add Notification button to add new email addresses to the notification list.
In addition to sending the RAS information collected by the Storage Automated Diagnostic Environment>, you can send a customizable subset of the event-driven messages from the Host (configured as Master) directly to local system administrators at the customer's site by email.
Note: Email might not be sent if the system is not properly configured to send mail to the recipient. This is primarily evident in Storage Service Processor environments,
where the Storage Service Processors are on a subnet, and there is no gateway to the intended recipient.
Send a Customizable Subset of Event-Driven Messages from the Host
| [ Up ] |
- Enter one or more email addresses into the Email Address/Filename textbox.
- For each address, choose from:
- Priority list:
- 0 = normal (green), information only. This is the lowest priority.
- 1 = warning (yellow)
- 2 = error (red)
- 3 = down (red), the system is down. This is the highest priority.
- Category list - Select from the current suite of storage products that are monitored by the Storage Automated Diagnostic Environment>.
- Event Type list - Select from a list of event types, categorized by device type.
- Click Add.
The Storage Automated Diagnostic Environment> sends the specific event type messages, by device type, to the email address(es) you specified.
For events that produce frequent error messages, specify the number of times you are notified.
If the maximum number is set to a value other than "no maximum" (the default), you can receive
an email 5, 10, 20, or 50 times within an eight-hour period.
Note: If you have set the maximum number of emails to 10, have received 10 emails on the
same component within an eight-hour time period, and want to receive more email notifications, you must click
the Clear Email maximums button.
Notification Providers
| [ Up ] |
Storage Automated Diagnostic Environment> providers encode data and interface with transport
mechanisms, which transmit information to the NSCC about configured storage
devices.
The Storage Automated Diagnostic Environment> supports seven providers: HTTP, Email, SSTR, NetConnect, SRS, SNMP Traps, and RSS. The provider's main function is to relay events generated by health monitors. The providers do this either by pulling the events from the Persistent Data Manager (PDM) or by registering them with the PDM for a certain type of event.
The types of providers are briefly described in the following sections:
- The HTTP Provider sends HTTP calls to an HTTP server and transfers CIM data in the appropriate format. This is an Intranet mechanism for transporting RAS information collected by the Storage Automated Diagnostic Environment agent and is currently internal to Sun only.
- The Email Provider enables specific events to be emailed to local administrators. Events can also be summarized and sent to a pager's email address.
- The SSTR Provider sends Storage Automated Diagnostic Environment> events to the Sun StorEdge Topology Reporter (SSTR) console. The SSTR console enables you to:
- View hardware assets such as hosts, host bus adapters, switches, and storage devices, including a graphical view of your environment
- Manage users, alarms, and assets
- Configure user email notification and network host notification
- Manage asset reporting (also known as Discovery)
- Launch other web browser-based applications related to devices in a particular environment
- NetConnect is a common information model (CIM) provider, which is an independent program. It requests information from the PDM, converts the information to the appropriate format, and relays it to NetConnect.
Note:
The Host with the Master agent must be the same Host that is configured as the relay in a NetConnect Provider configuration.
- The Sun Remote Services (SRS) Provider enables a second IP address, in case the events need to be sent to a second notification provider that also accepts the SRS XML format.
Note:
You must enter an IP address for the SRS Provider. A second IP address is optional.
- The SNMP Traps Provider enables the Storage Automated Diagnostic Environment> to send traps, for all actionable events that occur during monitoring, to external management systems.
When an alert occurs, it is sent to the SNMP transport as an SNMP trap. An SNMP trap listener can use the StorAgent.mib SNMP MIB file, which is included in the SUNWstade package to decode these alerts.
The alerts contain the following information:
- Storage Automated Diagnostic Environment Agent location
- Storage Automated Diagnostic Environment device to which the alert pertains
- Alert level
- Message content
- The RSS Provider enables the transfer of agent events to the RSS infrastructure using
UNIX-to-UNIX Communication Protocol (UUCP). RSS software is required on the host and is
intended to access the modem. It must be configured accordingly.
The RSS software is delivered with the Service Processor software bundle. The intended
audience for this software and for this provider are those customers who have purchased a
remote support service offering and are supplying phone lines to make use of the
modem phone home capability.
Note:
If the Storage Automated Diagnostic Environment> is run either manually or from the cron, and the RSS Provider is selected but the RSS software is not installed or is not configured properly, the following error message appears:
***ERR: Cannot find Machine name in Permissions file.
You will be prompted to push configurations whenever you change a configuration. The Review Configurations functionality lists
settings that you might have missed or you need to correct.
Note:
The Push Configs functionality is very important. It synchronizes all Slave instances under
a Master instance, with the instances of the Master to ensure proper monitoring and notification.
Start or Stop Agents
| [ Up ] |
You can control agent activity to temporarily stop the agent's cron
from running on a selected Host. You can also avoid creating email notifications on false errors when a device is being tested and faults are injected intentionally.
Note:
By default, the Master Monitoring function does not automatically default to On. You must manually enable Monitoring to On.
- Using the Start/Stop Agent(s) function to avoid creating email notifications on false errors works only if you are monitoring a single device type and all of the devices are monitored simultaneously, as in an upgrade.
- It would be wiser to use the Start/Stop Devices to control email notifications for specific devices.
Start/Stop Device Monitoring
| [ Up ] |
The Start/Stop Device Monitoring window enables you to deactivate monitoring on a device-by-device basis.
- To start or stop devices, check the appropriate checkbox.
- When Monitoring is 'on,' the 'Stop' checkbox is available.
- When Monitoring is 'off,' the 'Start' checkbox is available.
- Click Submit Changes. The Storage Automated Diagnostic Environment> must update the configuration files on all the Slave hosts configured with this Master.
- Use the [Push Configs] option on the Maintenance menu to accomplish this.
- If the device is being tested or if faults are being injected into the device
intentionally, you can temporarily stop the local notifications for a specific device
using the Start/Stop Device Monitoring window.
NOTE:
Monitoring continues when the device is deactivated (turned off). However, email
notifications will not occur for any faults that are detected while the device monitoring
is in this state. Consequently, any errors that may have been detected will be logged and
sent by means of the NetConnect or HTTP Providers, but not by means of email notification.
Use the Test Email window to send test emails and a message and to verify that the
mailing capability of the Storage Automated Diagnostic Environment> is installed and working properly.
To access the Test Email window:
- Click the Maintenance link in the Storage Automated Diagnostic Environment main window.
- Click the Test Email link on the Maintenance menu.
- Type your email address into the Email Address text box.
- Type a brief comment into the Message text field.
If you leave the Message text field blank, the test email contains a canned message with
the words Test Message in the subject line.
NOTE:
If you place a carriage return in the Message text field, you cannot enter text.
To restart, click in the Message text field and enter some text.
Review
Configuration | [ Up ] |
- Once you have completed your configuration and you want to verify all settings, select the Review Configuration link on the Maintenance menu.
- Follow the displayed instructions for settings that you might have missed or for those that you need to double-check.
Change Configuration Options
| [ Up ] |
The Configuration Options window returns the options depending on the location and status of the storage device. Use the following
information to update existing configuration options:
- Number of old (historical) topologies to save for comparison.
Type the number of topologies to save in Topology History for comparison. The default is 5.
- Test Manager Refresh Rate
Specify the window refresh (re-load) rate, in seconds, for test output. The default is 30 seconds, and the minimum refresh rate is 5 seconds.
- Max Event File Size
Specify, in megabytes, the maximum event file size. Used to limit data packet size to
notification providers. The default is 2.
- Path to Email Program
Use this field to change the local email program path.
- System Admin Email
Specify the email address of the system administrator responsible for receiving event output.
- Send One Email Per Event
Enable this checkbox to limit the number of emails sent about a specific event to one email.
- Maximum Number of Emails
Specify the number of emails that will be sent within an eight-hour period about a specific component. The default is no maximum; options include 5, 10, 20, or 50.
You can also clear the specified maximum number of emails using the
Email Notification window.
- GUI Bandwidth
Specify the bandwidth for low and high speed local area network (LAN). The default is Low Bandwidth.
- Create Hubs when appropriate
Enable this checkbox if you want Hubs to be drawn in the topology.
- Screen Size
Use to configure the default GUI screen size.