This chapter describes the diagnostic tests that can be run on Emulex adapters.
Use the Diagnostics tab to perform the following tasks:
n View PCI registers and Flash Contents information.
n Run the following tests on Emulex adapters installed in the system:
These tests are not available in read-only mode.
n Run a diagnostic dump and retrieve dump files from remote hosts (this option is not available in read-only mode).
n Control adapter beaconing (this option is not available in read-only mode).
ATTENTION: The Emulex HBA Manager application cannot retrieve data from an offline adapter or an offline port and will display incorrect information for that adapter or port. Adapters and ports must be online for the application to display accurate information.
Figure 63: Diagnostics Tab
The Diagnostics tab shows PCI register dump information (Figure 63). The information is read-only.
The Diagnostics tab enables you to run a quick diagnostics test on a selected port. The quick test consists of 50 PCI loopback test cycles and 50 internal loopback test cycles. This test is not available in read-only mode or on Emulex adapters in ESXi hosts.
ATTENTION: The Emulex HBA Manager application cannot retrieve data from an offline adapter or port and will display incorrect information for that adapter or port. Adapters and ports must be online for the application to display accurate information.
To run a quick test, perform the following steps:
1. From the discovery-tree (Figure 4), select the port on which you want to run the quick test.
2. Select the Diagnostics tab (Figure 63) and click Quick Test. A warning message appears (Figure 64).
Figure 64: Run Quick Test Popup
3. Click OK to run the test. The Quick Diagnostic Test window displays the test results.
The beaconing capability enables you to force a specific adapter’s LEDs to blink in a particular sequence. The blinking pattern acts as a beacon, making it easier to locate a specific adapter among racks of other adapters. This option is not available in read-only mode.
If you enable beaconing, the two LEDs blink rapidly in unison for 24 seconds, after which the LEDs report the adapter health status for 8 seconds. After the 8 seconds, the adapter returns to Beaconing mode. This cycle repeats indefinitely until you turn off beaconing or you reset the adapter.
ATTENTION: The Emulex HBA Manager application cannot retrieve data from an offline adapter or port and will display incorrect information for that adapter or port. Adapters and ports must be online for the application to display accurate information.
On supported adapters, you can also specify a specific beaconing duration, in seconds.
NOTE: The beaconing buttons are disabled if the selected adapter does not support beaconing.
To enable beaconing, perform the following steps:
1. From the discovery-tree (Figure 4), select the port whose LEDs you want to set.
2. Select the Diagnostics tab (Figure 63) and click Beacon On. The beacon status changes to On.
On supported adapters, you can also enter an optional duration time, in seconds, for the LEDs to blink. Enter the duration time and click Save.
To disable beaconing, perform the following steps:
1. From the discovery-tree (Figure 4), select the port whose LEDs you want to set.
2. Select the Diagnostics tab (Figure 63) and click Beacon Off. The beacon status changes to Off.
D_Port is a diagnostic mode supported by 16GFC and faster Brocade switches. D_Port tests enable you to detect physical cabling issues that result in increased error rates and intermittent behavior. If activated, D_Port runs a series of tests including local electrical loopback, loopback to the remote optics, loopback from the remote port to the local optics, and a full device loopback test with data integrity checks. It also provides an estimate of cable length to validate that a proper buffering scheme is in place. The various loopback tests allow some level of fault isolation so that you can distinguish faults from marginal cables, optics modules, and connector or optics seating. Bi-directional D_Port testing is supported.
NOTE: To initiate D_Port tests from the switch, Dynamic D_Port must be enabled on the adapter port. See Section 7.3.3, Enabling and Disabling Dynamic D_Port, for more information.
The following information applies to D_Port tests:
n Dynamic D_Port must be disabled on the adapter port to intiate D_Port testing using the Emulex HBA Manager application. See Section 7.3.3, Enabling and Disabling Dynamic D_Port, for more information.
n Dynamic D_Port must be enabled on the switch port.
n When in D_Port mode, the port is considered offline and can only initiate or respond to diagnostic tests (DPortTest).
n When using D_Port, the port enters D_Port mode and stays in that state until D_Port is disabled.
n D_Port testing is not available when FC port trunking (aggregation) is enabled.
n Basic connectivity diagnostics are already supported by Emulex HBAs. The Emulex HBA Manager application has diagnostic modes that support validation of the connection to the switch. The Brocade functionality provides the ability to diagnose marginal cable conditions (for example, dust in the optics) that result in higher error rates.
n D_Port tests run with the physical connection in an offline diagnostic state, so normal I/O cannot be sent through the physical port while the test is in progress. While the port is in D_Port mode, the link appears down on that port, similar to an unplugged cable.
n D_Port is also referred to as ClearLink®.
n For more information about D_Port, refer to the Brocade section of the Broadcom website at www.broadcom.com.
The D_Port Tests button on the Diagnostics tab (Figure 63) enables you to run D_Port tests from the Emulex HBA Manager application on supported adapters.
To run a D_Port test, perform the following steps:
1. From the discovery-tree (Figure 4), select the port on which you want to run the D_Port test.
2. Select the Diagnostics tab (Figure 63) and click D_Port Tests. The D_Port Tests dialog appears (Figure 65).
3. Click Start Tests. The start time is displayed.
Figure 65: D_Port Tests Dialog
The following D_Port Tests dialog fields are displayed:
– Overall Test Result – PASSED, FAILED, or a specific error, depending upon the outcome of all the test phases.
– Frame Size – The size of the frames used in each test phase.
– Buffers Required – The number of buffers required for each test phase.
– Roundtrip Link Latency – Estimated roundtrip link latency calculated by the switch during the execution of all tests.
– Estimated Cable Length – Estimated cable length calculated by the switch during the execution of all tests.
– Test Phase – The name of the test run.
– Phase Result – The result of the test run. Possible results are Passed, Failed, or Skipped.
– Phase Latency – The round trip legacy (in ns) calculated during the execution of the test.
– Local Error – The errors, if any, detected on the local side of the test.
– Remote Error – The errors, if any, detected on the remote side of the test.
To stop running D_Port tests, click Stop Tests. The stop time is displayed.
To save the test results to a file, click Save Results to File. You are prompted to enter the file name.
NOTE: If the SFP or adapter firmware does not support running D_Port diagnostics, clicking Start Tests causes an error message to be displayed, and the tests are not run.
The Emulex HBA Manager application allows you to trace the communication route for FC packets transmitted between an FC initiator port and an FC target port.
Communication route information, such as the switch name, domain ID, ingress and egress port name, and ingress and egress physical port number, is accumulated for all switch ports through which packets are routed. Data is collected for both the outward-bound route from the initiator to the target and the inbound route from the target to the initiator.
The FC Trace Route button on the Diagnostics tab enables you collect fabric routing information (Figure 66).
The following considerations apply to FC trace route:
n FC trace route must be supported by the fabric.
n FC trace route support is provided on Windows, Linux, and ESXi operating system platforms.
Figure 66: Diagnostics Tab (FC Trace Route Button)
To enable FC trace route, perform the following steps:
1. From the discovery-tree (Figure 4), select the port on which you want to enable FC trace route.
2. Select the Diagnostics tab (Figure 66) and click FC Trace Route. The FC Trace Route dialog appears (Figure 67).
Figure 67: FC Trace Route Dialog
3. The Target Port list displays the WWPNs of all targets that are seen by the initiator port. Select a target port and click Start.
The FC Trace Route dialog displays fabric routing information between the selected initiator and target ports (Figure 68).
NOTE: If an error occurs when processing the FC trace route request, a message is displayed at the bottom of the dialog.
Figure 68: FC Trace Route Dialog with Route Information Displayed
The following information is collected for each trace route:
n Switch Name – The switch chassis WWN.
n Domain ID – A number used to uniquely identify a switch in a fabric. This number is assigned by a fabric administrator as part of fabric configuration. The domain ID is an 8-bit field whose value ranges from 0 to 255.
n Ingress Port Name – The WWN of the physical port through which an FC packet enters a specific switch.
n Ingress Port Num – The physical port number of the port through which an FC packet enters a specific switch.
n Egress Port Name – The WWN of the physical port through which an FC packet exits a specific switch.
n Egress Port Num – The physical port number of the port through which an FC packet exits a specific switch.
Click Stop to stop the trace route request.
Click Save to File to save the results of the most recent FC trace route operation to a log text file. The default file name for the log text file is FCTraceRte_IWWPN_TWWPN (where IWWPN is the initiator WWPN and TWWPN is the target WWPN. You can change the file name.
Click Clear to erase the FC Trace Route Table information from the dialog.
On Windows and Linux systems, you can manage firmware logs using the Diagnostics tab.
The FwLogBufferCnt parameter on Windows systems and the ras-fwlog-buffsize parameter on Linux systems must be enabled to create firmware logs. Refer to the Emulex Drivers for Windows User Guide or the Emulex Drivers for Linux User Guide for instructions on enabling the parameters.
NOTE: After you enable firmware error logging and reboot the system, error logging starts automatically with the log level set to 0.
Using the Emulex HBA Manager application, you can interact with firmware logs in the following ways:
n Retrieve an existing firmware log and save it to a file in the dump directory.
To manage firmware logging, perform the following steps:
1. From the discovery-tree (Figure 4), select a port whose firmware information you want to retrieve.
2. Select the Diagnostics tab (Figure 63) and click Firmware Log. The Firmware Logging dialog appears (Figure 69).
Figure 69: Firmware Logging Dialog
To stop firmware logging, select Stop.
To dump a copy of the log onto the C:\Program Files\Emulex\Util\Dump directory, click Retrieve Log. (You do not need to stop logging first.)
To change the log level of a firmware log, select Stop, and then select the desired log level. Available choices range from 0 (Least Verbose) to 4 (Most Verbose).
To resume logging after it is stopped, click Start.
To close the Firmware Logging dialog, click Exit.
10.7 Creating Diagnostic Dumps
Diagnostic dump enables you to create and manage a diagnostic dump for a selected adapter. Dump files contain information such as firmware version and driver version, which is particularly useful when troubleshooting an adapter.
You can retrieve user-initiated and driver-initiated driver dump files, delete the dump files, or repeat the process on all resident dump files. You can also retrieve or delete dump files from remote hosts.
NOTE: This option is not available in read-only mode.
CAUTION! Disruption of service can occur if a diagnostic dump is run during I/O activity.
NOTE: For VMware systems, you must set a dump directory before initiating a dump. The dump directory must be a storage partition (a datastore) under the directory /vmfs/volumes.
To start a diagnostic dump, perform the following steps:
1. From the discovery-tree (Figure 4), select an adapter whose diagnostic information you want to dump.
2. Select the Firmware tab (Figure 23), and click Diagnostic Dump. The Diagnostic Dump dialog appears (Figure 70).
For hosts being managed through the CIM interface, the Set Dump Directory button enables you to set the dump directory for ESXi host dumps (VMware only).
3. Specify how many files that you want to retain by selecting a number from the Dumps Retained list and clicking Apply.
4. Click Delete Existing Dump Files to remove existing dump files for the selected adapter from your system.
Figure 70: Diagnostic Dump Dialog
5. Click Start Dump. Dump files are created. The file location depends upon your operating system:
– Windows – %ProgramFiles%Util\Dump\.
– Linux – /var/log/emulex/ocmanager/Dump.
– VMware – A dump directory that you create under /vmfs/volumes.
l <Hostname_WWPN_Date-Time>.efd
l <Hostname_WWPN_Date-Time>.txt
6. To obtain remote host dump files and copy them to your local system, click Get Dump Files. The Diagnostic Dump File Transfer dialog appears (Figure 71).
NOTE: The Get Dump Files button is dimmed if a local adapter port is selected.
Figure 71: Diagnostic Dump File Transfer Dialog
7. Select the files you want to copy (multiple selections are available), and click Start Copy. The remote dump files are copied to your local dump folder. The local dump folder locations are described in Step 5.
10.8 Running Advanced Diagnostic Tests
The advanced diagnostics capability gives you greater control than the quick test over the type of diagnostics tests that run. Through advanced diagnostics, you can specify which tests to run, the number of cycles to run, and the actions to take in the event of a test failure. Advanced diagnostics capability is not available in read-only mode.
To run advanced diagnostics tests, click Advanced Diagnostic Tests on the Diagnostics tab (Figure 63). The Diagnostic Test Setup dialog appears (Figure 72).
Figure 72: Diagnostic Test Setup Dialog
The following Diagnostic Test Setup dialog fields are displayed:
– External Loopback (requires loopback plug)
NOTE: For details about these tests, see Section 10.8.1, Running Loopback Tests.
n End-to-End (Echo) Test area:
NOTE: For details about this test, see Section 10.8.2, Running End-to-End Tests.
NOTE: You cannot run the external loopback test and the echo test concurrently. If you select External Loopback, Echo Test is dimmed, and if you select Echo Test, External Loopback is dimmed.
This area enables you to define the actions to be performed in the event of a test failure.
– Stop Test – Do not log the error, and halt the test. No further tests are run.
– Ignore – Log the error, and proceed with the next test cycle.
This area enables you to specify test cycles in three ways:
– Select an established cycle count by choosing the desired number.
– Enter a custom cycle count in the blank field.
– Set the test to run until you click Stop Test by selecting Infinite.
Enter a custom test pattern to be used in tests that transfer data. The test pattern can be up to 8 hexadecimal bytes.
This area displays the number of completed cycles of each test run, in addition to the number of errors for each test.
For details about test logs, see Section 10.8.3, Saving the Log File.
You can run the following loopback tests:
n PCI loopback test – A firmware-controlled diagnostic test in which a random data pattern is routed through the PCI bus without being sent to an adapter link port. The returned data is subsequently validated for integrity.
n Internal loopback test – A diagnostic test in which a random data pattern is sent to an adapter link port and is then immediately returned without actually going out on the port. The returned data is subsequently validated for integrity.
n External loopback test – A diagnostic test in which a random data pattern is sent to an adapter link port. The data goes out the port and immediately returns by way of a loopback connector. The returned data is subsequently validated for integrity.
The following information applies to loopback tests:
n You cannot run the external loopback test and echo test concurrently. If you select External Loopback, Echo Test is dimmed, and if you select Echo Test, External Loopback is dimmed.
n Adapters and port information are not available during diagnostic loopback tests.
n Internal and external loopback tests on trunking-enabled ports do not support infinite test cycles.
n Internal and external loopback test results are displayed for each physical port.
n Each physical port must have a loopback connector when performing external loopback tests on trunking-enabled ports.
To run loopback tests, perform the following steps:
1. From the Diagnostics tab (Figure 63), click Advanced Diagnostics Tests (Figure 72). From the Loopback Tests area of the dialog, choose the type of loopback test that you want to run, and specify the error action, the test cycles, and test pattern.
NOTE: You must insert a loopback plug in the selected port before running an external loopback test.
2. Click Start Test. The following popup appears (Figure 73).
Figure 73: Run Diagnostic Tests Popup
3. Click OK. If you chose to run an external loopback test, the following popup appears (Figure 74).
Figure 74: Run Diagnostic Tests Popup for External Loopback
4. Click OK. A progress bar indicates that the test is running.
Periodic test feedback, consisting of the current loopback test/cycle plus the completion status of each type of test, is displayed in the Test Log area of the dialog. Either click Clear to erase the contents of the log display, or click Save to file to save the log file.
After starting the tests, you can click Stop Test to stop the tests before they complete. Depending upon the tests being run, it might take some time before they stop.
10.8.2 Running End-to-End Tests
The end-to-end test, or echo test, enables you to send an echo command and response sequence between an adapter port and a target port.
NOTE: Not all remote devices respond to an echo command.
NOTE: You cannot run the echo test and the external loopback test concurrently. If you select Echo Test, External Loopback is dimmed.
To run echo tests, perform the following steps:
1. From the Diagnostics tab (Figure 63), click Advanced Diagnostic Tests (Figure 72).
2. Select Echo Test, and enter the WWPN for the target.
a. Click Select from list if you do not know the WWPN of the test target. The Select Echo Test Target dialog appears (Figure 75).
b. Select the port to test from the tree-view. All relevant information for the selected port is automatically added to the Targets area of the dialog.
Figure 75: Select Echo Test Target Dialog
3. Select the desired target from the Targets list, and click Select. The Diagnostics dialog reappears.
4. Define the other parameters that you want to use and click Start Test. The following warning popup appears (Figure 76).
Figure 76: Run Diagnostic Tests Popup
5. Click OK. A result screen appears, and the test results appear in the test log. Either click Clear to erase the contents of the log display, or click Save to file to save the log file.
You can save the test log to a file for later viewing or printing. When new data is written to a saved file, the data is appended to the end of the file. Each entry has a two-line header that contains the identifier of the adapter being tested and the date and time of the test. Over time, the data accumulates to form a chronological history of the diagnostics performed on the adapter.
The default file location is as follows:
n In Windows: The Emulex HBA Manager application installation directory on your local drive.
n In Linux: /var/opt/emulex/ocmanager/Dump
n In the VMware server: A default directory does not exist for VMware.
After writing an entry into the log, you are prompted to clear the display. The default name of the saved file is Diagtest.log. An example of a saved log file is shown in Figure 77.
Figure 77: Example of a Diagtest.log Window
To save the log file, perform the following steps:
1. After running a test from the Diagnostic Test Setup dialog (Figure 72), click Save to file. The Select Diagnostic Log File Name dialog appears. The default name of a saved file is DiagTest.log.
2. Browse to the desired directory, change the log file name if you want, and click Save.