This section describes the maintenance action plans (MAPs) for the SNC. Maps
exist for all of the SNC's component systems. If it is clear that a particular
component system is at fault in a problem situation, you may go directly to
the MAP for that component by consulting Table 1 below.
Otherwise, you may choose to work through the Events and Actions,
in Tables 2 - 5 below. You may also choose to work through the Start
MAP in order to determine which MAP is appropriate for your situation.
Table 1. Maps Troubleshooting Table
Component
Subsystem
Go
To
If
you are having problems with your system, but are unsure which subsystem
to troubleshoot first
The event log is about to overwrite the earliest events
No action necessary.
(1)
Check the event log to find out which interface (FCAL or SCSI) caused this event.
Then use the appropriate action number. (2) This
message signals a change in state. A string sent with the message will indicate
the nature of the previous state. (3) Events not logged in SNC event log.
(1) Events not logged in SNC event log. (2) Not reported. View the server log. (3) Event reported only when SNC Manager 2.7 or above is running
Start MAP
Always start problem determination here and review the following steps. Gather
as much information as possible before performing a repair action. When gathering
information, you may need to connect the service terminal to the service port.
See Connecting to the Service Port.
For the latest information on the SNC, see the website and technical support
contact information at support.dell.com.
Event Code or Obvious Symptom
An error has been observed. If the event code or error symptoms are known,
go to the Events and Actions with that event code or symptom,
and perform the recommended action.
Visually Inspect LEDs
Observe front-panel SNC LED status indicators and check against the Events
and Actions. For more information about expected LED behavior, see LED
Indicators.
If the RDY LED is flashing as expected but the SCSI, Ethernet and FC connection
LEDs are off, the SNC may have been left in diagnostic mode. The device drivers
for SCSI, FC and Ethernet interfaces are disabled when in diagnostic mode. If
the command prompt on the service terminal is diagmode
> , return the SNC to normal mode by entering the normalBoot
command on the service terminal. See Boot
Modes.
Check for Problems on Attached SCSI Devices
Check the following on the SCSI devices to determine whether they are the
source of the problem:
LEDs
Display panels
Firmware levels
Operability
Check FC Host Versions
For an updated list of supported SNC host platforms and FC host bus adapters,
see the website and technical support contact information at support.dell.com.
Determine the following:
Operating system version
Service pack version
Hot-fix version
HBA hardware version
HBA firmware version
HBA device driver version
If an update is required, perform the update.
Check Dell SNC Product Versions
Hardware version information is viewed by using the sysVpdShow
command from the service terminal.
If your service representative determines that the hardware version is at issue
in your situation, you may need to use the sysVpdShow
or sysVpdShowAll command.
When viewing the log, set the viewing level to warning. The following message
should be in the list:
000001 0185 0d:00h:00m:05s:15t -- NOTICE: LOGGING STARTED
If not, then continue viewing the event log until the start message can be
found in the log. If this start message cannot be found, then use the log message
with the lowest sequence number (the first number on the display line) as the
start of the current log.
If the client application is not available, use the loggerDumpCurrent
2 command from the service terminal. See the loggerDump
[number] command. Check the event codes against the Events
and Actions. To see additional log messages, follow the procedure Event
Log Dump.
Quick Component Check
From the service terminal, enter the showBox
command. See showBox. If an installed component
does not show up in the display, go to the correct MAP. For example, if a SCSI
interface is missing, go to SCSI MAP. If a FC is missing,
go to FC MAP.
Check the Host Event Log
Check the event log on the FC host. Look for the most recent entries and check
to see if there are any FC HBA driver errors. If there are, go to FC
MAP.
Review all visual observations and event codes. Where multiple action plans
are indicated in the Action column, more than one action may be required
to determine the problem. Refer to the Events and Actions
section and carry out the action plan linked to the table entry.
Table 2 shows a description of the visual observations
of events with no codes. The action plans specify steps you should should take
to resolve a problem before you visit support.dell.com.
When multiple action plans are indicated in the Action column, more than
one action may be required to determine the problem.
Database Full MAP
These steps are performed if event code 42 is
reported indicating the persistent address map database is full (more than 255
devices are in the database). If the database has no room for a newly detected
device, the new device will not be mapped (assigned a LUN). The database can
become full if there are more devices attached than there were previously or
if devices were moved to different ports or channels. You can free up database
entries that are no longer needed, but keep the devices that are currently attached
at the same assigned LUNs.
NOTE: Use this procedure only
when you are sure that the devices you are interested in are connected and
available to the SNC. Devices that are not currently attached will be removed
from the database. You must reboot the SNC after performing this procedure
for the changes to take effect.
Press the <Enter> key on the service terminal. If the prompt is not
displayed, go to Service Port MAP to determine whether
the RS-232 cable and service terminal are working properly.
From the service terminal, enter the mapShowDatabase
command to display the contents of the database.
If the library is off, then turn it on and wait until the RDY LED is blinking
once per second.
As the SNC is booting, several status messages should be displayed on the
service terminal.
If no status messages are displayed, go to Service Port
MAP to determine whether the RS-232 cable and service terminal are working
properly. Otherwise, look for the last status message Done
executing startup script.
Enter the showBox command from the service
terminal.
NOTE: The SNC supports up to 256
LUNs. If more than 256 LUNs are attached this will result in unreliable
behavior. Only 255 LUNs are available to the user, for the SNC takes up
one LUN (0) as its Command and Control LUN.
Issue the scsiShow command from the service
terminal to display a list of attached SCSI devices.
Only the Dell PowerVault 136T tape library and tape drives, which are available
with the library, are supported. Any devices on the list that are not supported
must be removed from the SNC.
If the attached SCSI devices are all supported devices, go to SCSI
Loop Back Test.
Check SCSI Bus Termination
NOTICE: If in this procedure you
determine that you need to remove a SCSI cable or terminator, all I/O to
the SNC must be stopped and the library must be turned off.
For each SCSI device attached to the questionable SCSI channel:
Verify the termination.
The SNC has internal terminators on its SCSI channels.
NOTICE: If in this procedure you
determine that you need to correct any SCSI IDs, all I/O to the SNC must
be stopped. If two or more devices on the same SCSI channel are configured
at the same SCSI ID, only one of those devices will be seen by the SNC.
Data transfers to that device will be unreliable.
Write down the SCSI IDs of all devices connected to each SCSI channel.
Check that only one device is set to each ID for each SCSI channel.
If a target device is set to the same ID as the SNC (ID 7), the bus will
be unstable and data corruption may result.
Check that for each SCSI Channel, no devices are set to ID 7 (except the
SNC's channel ID).
NOTICE: Only LVD devices must be attached
to LVD SCSI channels.
NOTICE: If in this
procedure you determine that you need to replace a SCSI device, all I/O
to the SNC must be stopped and the library must be turned off.
When a mix of Ultra2/3 SCSI and Ultra SCSI devices are connected to a single
bus, the bus will run at the Ultra SCSI speed.
Because the bus will auto-adjust to the slowest speed, it is not recommended
that you run both Ultra2/3 SCSI and Ultra SCSI devices on the same bus.
If you do not have to replace an improper device, go to Examine
SCSI Cables.
Replace the improper SCSI device with a proper device.
Reconnect the SCSI cable.
Turn on the library.
After the SNC finishes booting, enter the scsiShow
command from the service terminal.
Compare the list of attached devices with the physical devices.
NOTICE: If in this procedure you determine
that you need to replace a SCSI device, all I/O to the SNC must be stopped
and the library must be turned off.
Look for damaged cables.
Check for breaks in the cable jacket, exposed or frayed cable shield,
exposed or broken wires.
Replace any damaged cables.
Look for inadequate cables.
Older SCSI cables may not be suitable for running at Ultra2/3 speeds.
Be sure all cables are rated for Ultra2/3.
Replace any that are not.
Check for mixed cable types.
If a SCSI bus has both round cables and flat ribbon cables, it may suffer
problems when running at Ultra2/3 speeds.
Use the same cable type consistently for all segments of the SCSI bus.
Replace any cables that do not match.
Check for unshielded cables.
An unshielded SCSI cable
used external to a cabinet may cause reliability problems, due to interference
from other electrical devices.
NOTICE: Before performing diagnostics, the
SCSI cables must be disconnected from the SNC. Before changing any SCSI
cables, all I/O to the SNC must be stopped and the library must be turned
off. Do not perform SCSI loop back tests on SCSI channels that are attached
to SCSI target devices. If you do this, you will corrupt the data on the
device.
If the SNC's command prompt on the terminal is not diagmode
> , enter the diagBoot
command on the service terminal.
Remove the cable from one attached
SCSI device, and attach it as a loop back cable to the SNC.
Power on the library.
When the SNC finishes booting, enter the scsiChannelTest
x, y command from the service terminal (where `x' and `y' are the slot
numbers connected by the loop back cable) to perform a loop back test.
If the SNC returns a
FAILED status, the cable is faulty and must be
replaced.
If the SNC returns a
PASSED status, go to Isolating
SCSI Devices and try to isolate bad devices on the SCSI bus.
Attach ONLY the device in question
to the SCSI channel it was originally connected to, using a known-good SCSI
cable and terminators. Refer to the list made in Check
for Multiple SCSI IDs.
Power on the library and wait
for it and the SNC to finish booting. The Library takes longer to boot
than the SNC.
Enter the scsiShow command from the service
terminal and verify the device's presence.
If the device is missing, then it needs to be replaced and/or serviced.
Inform the system administrator
of any suspected bad devices found in this step.
Replace or remove any bad
devices.
If the device is present,
then perform a health check.
With the known-good GBIC installed, confirm that the SNC's FC connection
is correctly displayed by issuing the showBox
command from the service terminal.
If showBox
displays the FC correctly, exit this MAP.
If the output of the showBox command still
does not match the FC configuration, remove the known-good GBIC and reinstall
the original GBIC.
The optical cable diameter is important for long wave FC optical transceivers.
Short wave GBICs will function with any cable diameter (though not necessarily
all lengths). The cable diameter is expressed as the ratio of the diameter of
its fiber core over the diameter of the protective cladding.
Write down the core/cladding
diameter printed on the cable jacket. Also write what FC port it was plugged
into.
For each FC that is listed
as FCOLW, check the core/cladding diameter on the jacket. Only cables with
9/125 diameter are acceptable.
Replace any 50/125 or 62.5/125
cables with the 9/125 rating, then retest.
Remove the loop back plug and
reconnect the FC cables.
Replace
FC Cable
If the library is still on, turn it off.
Replace the cables to the
FC ports with known good cables.
Power up the library and wait
for it and the SNC to finish booting. The library takes longer to boot than
the SNC.
If problems persist, replace
other external devices that are attached to the FC ports (go to
Replace FC Device).
Otherwise, exit this MAP.
Replace
FC Device
The device that needs to be replaced
is one of the following:
FC HBA
FC switch
FC hub
FC disk subsystem
Inform the system administrator
that it is necessary to replace the FC device connected to the SNC to resolve the FC errors. Repeat this MAP after replacing
an external component.
SNC MAP
The following steps are performed
if:
RDY LED not blinking once per second after power has been on for one minute
The SNC is not responding
SNC processor memory
fault detected
SNC PCI bus parity
error detected
SNC PCI interface
error detected
The Server could not verify
the connection to the SNC
Observe Operational LED Behavior
When the library is first powered
on, the front-panel SNC's LEDs flash a variety of patterns as it performs the
power on self test (POST) and then starts booting. Within one minute, the SNC
should have booted successfully and the RDY LED should be blinking once per
second. If the RDY LED is blinking as expected, go to Start
MAP. If not, perform the following step:
See the envMonShow
command for operational temperature ranges that will generate these messages.
If the temperature problems
are due to a reduction in blower functionality, event codes 62, 63, 64, or 65
will be accompanied by event codes 67 or 68.
If room temperature is within
satisfactory limits, and the SNC is generating event codes 62-65, or 67 then
check air outlets on the SNC connector panel for obstructions.
If any are found, remove them
and permit the SNC to cool.
If temperatures stabilize
(trap event codes 61 or 66 are received) exit this MAP.
If you were directed to this
MAP because of trap event code 67, replace the SNC at the next scheduled maintenance.
If no obstructions are found,
or temperatures in the SNC do not stabilize to satisfactory limits within
10-15 minutes, or if the SNC continues to generate event code 62, perform
a graceful shutdown of the library and consult library service personnel.
See Getting Help section.
The network administrator must
provide the following information before you can perform this MAP:
The IP address for the SNC
- REQUIRED
The net mask for the SNC
in decimal and hex formats - REQUIRED
The network gateway IP address
for the SNC - if assigned
The IP address of a computer
on the same subnet as the SNC for PING tests - REQUIRED
The IP address of the Dell
PowerVault SNC Bridge Manager Server - REQUIRED
The SNC must be booted and its Ethernet port must be attached to the local
area network. See Figure 2.
Figure 2. Ethernet port on SNC Face Plate
Verify that the Ethernet Link LED, which is the green LED on the left,
is on. This LED is a combined link/activity LED. If the link is up and there
is activity, it will blink.
Remove the Ethernet cable from the SNC Ethernet port
and attach an Ethernet loop back plug.
Verify that the Ethernet LEDs are functioning correctly. If they are not
functioning correctly, inform the system administrator that there is a network
hardware problem where the SNC is attached.
Remove the Ethernet loop back plug.
Obtain another Ethernet cable. Use this cable to attach the SNC to the
local area network.
If the prompt is not displayed, go to Service Port MAP
to determine whether the RS-232 cable and service terminal are working properly.
From the service terminal, enter the ifShow
command.
SNC > ifShow
lnPci (unit number 0):
Flags: (0x63) UP BROADCAST ARP RUNNING
Internet address: 192.168.1.56
Broadcast address: 192.168.1.255
Netmask 0xffffff00 Subnetmask 0xffffff00
Ethernet address is 00:60:45:0d:00:c1
Metric is 0 Maximum Transfer Unit size is 1500
2 packets received; 4 packets sent
0 input errors; 0 output errors
0 collisions
lnPci (unit number 0):
Flags: (0x69) UP LOOPBACK ARP RUNNING
Internet address: 127.0.0.1
Netmask 0xff000000 Subnetmask 0xff000000
Metric is 0
Maximum Transfer Unit size is 4096
4 packets received; 4 packets sent
0 input errors; 0 output errors
0 collisions
From the ifShow display, write down the values of the Internet address,
netmask, and subnetmask.
Compare the Internet address with the IP address supplied by the network
administrator.
Compare the netmask value with the netmask in hex format supplied by the
network administrator. The subnetmask should be the same as the netmask. If
these values are correct, go to step 20.
Otherwise, use the ethAddrSet command to set
the correct IP address and netmask values See the ethAddrSet
command.
From the service terminal, enter the reboot
command and wait for the SNC to finish booting.
Enter the command ping <host
IP address>, 10 where <host IP address> is four decimal numbers
separated by periods. This is the address provided by the network administrator
for PING testing.
Display of successful PING test
SNC > ping 192.168.1.1, 10
PING 192.168.1.1: 56 data bytes
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
64 bytes from 192.168.1.1: icmp_seq=0. time=0. ms
----192.168.1.1 PING Statistics----
10 packets transmitted, 10 packets received, 0% packet loss
round-trip (ms) min/avg/max = 0/0/0
value = 0 = 0x0
Display of failed PING test
SNC > ping 192.168.1.251,10
PING 192.168.1.251: 56 data bytes
no answer from 192.168.1.251
value = -1 = 0xfffffff
If the test passed, remove the Ethernet loop back plug.
From the service terminal, enter the normalBoot
command and wait for the SNC to finish booting.
Reattach the Ethernet cable to the SNC.
From the service terminal, enter the gateAddrGet
command and write down the network gateway address that is displayed.
Compare this address to the one provided by the network administrator.
If both addresses are the same, go to step 34.
SNC > gateAddrGet Gateway Address set to 192.168.1.1 value = 0 = 0x0
If the network gateway address is incorrect, set it to the value provided
by the network administrator using the gateAddrSet
command.
From the service terminal, enter the reboot
command and wait until the SNC has finished booting.
From the service terminal, use the ping command
to PING the IP address of the Dell PowerVault SNC Bridge Manager Server. See
step 20.
From the Dell PowerVault SNC Manager Server, PING
the SNC.
If both PING tests succeeded, exit this MAP.
If the PING test fails, inform the network administrator to check and correct
the network connection, route tables, and network gateway addresses for both
the Dell PowerVault SNC Manager and the SNC.
Service Port
MAP
Check the RS-232 Cable
This test requires another laptop
or desktop computer with a functioning RS-232 9-pin port. Terminal emulation
software must be installed and running.
Remove the RS-232 cable from
the service port, and connect it to the compatible port on the other computer.
Connect the service terminal
to the cable.
Set the service terminal and
the other computer to 19200 baud, 8 data bits, no parity, one stop, XON-XOFF.
Enter test lines in the service
terminal and the other computer. Confirm that they are displayed on each other.
If this fails, replace the RS-232 cable.
If the cable is good, remove it from the other computer.
*Offers subject to change. Taxes, shipping, handling and other fees apply. U.S. Dell Small Business new purchases only. LIMIT 5 DISCOUNTED OR PROMOTIONAL ITEMS PER CUSTOMER. LIMIT 5 VOSTRO OR INSPIRON UNITS PER CUSTOMER. Dell reserves right to cancel orders arising from pricing or other errors.
