Troubleshooting of Brocade SAN Switch

               

          Basic Troubleshooting of Brocade SAN Switch

1. Login Procedure for the SAN switch

Please follow the below steps whenever you suspect issue with Switch ports:

1. login jumpstation using jumpstation accounts
2. Login to Storage management server (SMA)
3. Telnet to switch with admin privileges (Eg: user name=admin; Password=password)

2. Checking the port status on the SAN switch

a.     Once logged into the switch with admin privileges, execute the command Switchshow

b.   The output of the command will be as shown in the following screenshot

c.   The status of the port will be either of the following states as shown in the next page

4. Port should be online (Either F port or E port) and if the port is not online, it will be having one of the statuses mentioned above. Please make a note of necessary actions that have to be taken for the following most common states:

No_Module: There is no SFP installed and hence this port needs SFP to accept logins from any devices connected to this port

No_Light: There is no light coming from the node. This means there is some physical problem like cable, SFP, even HBA might be faulty

No_Sync: This means there is light and it is not physical problem. This is something to do with configuration on the switch (like port speed not matching)

In_Sync: This means the node HBA and switch ports are in synchronization but there could be driver issue on the Host HBA. That is why the port is not coming online

Laser_flt: This means defective GBIC/SFP. Replace them 

Port_Flt: This means port has been marked faulty (defective GBIC/SFP, Cable or device)

Online: The healthy port is always online

3. Checking the port for CRC errors

1. After login to switch console, execute the command Porterrshow to know the crc errors for the ports

2.  The output of the command will be as shown in the following screenshot

3. If the there are more CRC errors, we may have to replace the SFP. But before replacing SFP clear the port statistics by executing the following command

      portstatsclear <slot number/port number>

4. After this, monitor the port for 24 hours and if CRC errors go on increasing, replace the SFP.

4. Checking whether switch port went offline

1. After login to switch console, execute the command fabstateshow to know whether the port went offline/online as shown in the below screenshot

2. In the below example, the port 7 went offline and came back online

5. Checking the switch error log

1. After login to switch console, execute the command errorshow to see any error messages in the switch log

2. Please go on entering till the log for the required date is displayed

 6. How to find out physical location of SAN device in the fabric

1. After login to switch console, execute the command nodefind to know the physical location of the device in the fabric, alias of the device and many more details of the device

2. In the above screenshot, the red mark gives the FC address of the port (PID)

3. Make note of PID which will give the location of the device. In the above screenshot, it is 292c00. This is hexadecimal number where first two digits gives the domain id, next two digits give the port number and last two digits are always zero in case there is no loop device in the fabric. For more details please refer to appendix

4. Yellow mark gives the type of HBA, model of HBA, firmware version of the HBA. In this example, this Emulex HBA model – LP1050 with Firmware version (1.91A5)

5. Green colour gives the port number where node is connected

6. Blue colour gives the alias of the node in the zone

7. How to find out Domain id of all the switches in the fabric

1. After login to switch console, execute the command fabricshow to know the domain id of all the switches in the fabric as shown in the below screenshot

2. In the above screenshot, the first column marked with red colour gives the domain id of all the switches in the fabric

3. Yellow mark where there is mark (>) indicates the principal switch in the fabric

 8. How to collect switch logs to be sent to HPRC

1. Login to Storage management server (SMA)
2. Open putty by clicking the icon .
3. The Putty program will open as shown in the screen shot below

d.Click on session and mention the IP address of the switch & connection type (By default SSH) screenshot below:

5. Click on Logging and select all session output as shown below:

6. Click on browse to specify the file name and its path where the output is dumped
7. Click on open to open the console
8. Specify the user name and password

9. Once logged into switch console, execute the command supportshow
10. After this command is completely executed, exit the console of the switch by entering the command exit
11. Find the output file generated from putty and send it to HPRC for further analysis

9. Appendix

This appendix covers the concept of addressing mode used in SAN.

PID FORMATS

THE FORMAT OF A 24-BIT ADDRESS IN NATIVE MODE

XX1YZZ

XX is a value between 0x1 to 0xef inclusive (Domain ID 1-239 in decimal)

The “1” means Native Mode

Y is the port number 0x0 to 0xf (0-15 decimal)

ZZ is the AL_PA for a loop device or 00 for an F_Port

THE FORMAT OF A 24-BIT ADDRESS IN CORE PID MODE

XXYYZZ

XX is a value between 0x1 to 0xef inclusive (Domain ID 1-239 in decimal)

YY is the port area

ZZ is the AL_PA for a loop device or 00 for an F_Port

THE FORMAT OF A 24-BIT ADDRESS IN EXTENDED EDGE PID MODE

XXYYZZ

XX is a value between 0x1 to 0xef inclusive (Domain ID 1-239 in decimal)

YY is the port area + 0x10, wrapping at 0x7f

ZZ is the AL_PA for a loop device or 00 for an F_Port