Blade center s chassis from scratch
1. DVD and quick start install guide
2. Two people to lift out of box
3. 7 u tall 6 blade bays disk storage modules
4. Battery backup
5. Put DVD in Thinkpad
6. It will auto start
7. Plug power into chassis
8. License agreement
9. There's a learn more link to tell you about the unit
10. Click get help first to see where to plug it in
11. Chassis connection choose direct connection
12. Next
13. Leave default
14. Next
15. Choose default IP
16. Next
17. 1 or 2 of the components have problems
18. Attention alert on Ethernet and SAS modules
19. It’s just a firmware issue so click next
20. Click default "express"
21. Next
22. IP Configuration next
23. Firmware update to AMM and SAS next
24. Go with default date and time
25. Do not wish to change default password
26. You can put event notification if you want or click next
27. Storage zone configuration
28. Note at first all drive bays and disks have been assigned to blade
29. Choose assign 2 disk drive bays to each blade
30. You'll have redundancy for the OS
31. Click begin configuration after checking details
32. Wait a long time and click next
33. Finish and done
34. Now you can apply the OS
DS4000 fabric storage manager
1. We can create the volume from a new array from the currently un-configured capacity
2. Right click on the total un-configured capacity
3. Select "create logical drive"
4. Select desired RAID level 1, 3, 5, or 10
5. Even though 10 is not in the list you can create one by selecting 4 or more disks and select RAID 1
6. But since this will be used for a RAID storage volume we will select RAID 5
7. Then choose the drives that your RAID 5 array will consist of
8. This can be done manually allowing for the selection of exact drives in the drive drawer
9. You can choose enclosure loss protection which allows for the loss of an entire drive enclosure without disruption to the array
10. Create an array with 4 drives and enclosure loss protection
11. Select parameters for the logical drive
12. Create a 50GB drive
13. Name it DB2 Data
14. Select "custom settings" to further tailor this drive to your database application
15. Select "database" for the type
16. Cache read ahead changes to 0 and the stripe size to 128
17. If your database recommends a stripe size if 64 you can choose "custom"
18. You can now assign the LUN to the default group or assign it later
19. Choose to assign it later and click finish
20. In left frame of GUI you now see your new RAID 5 array and your 50GB LUN
21. 1 drive in each enclosure is utilized giving you enclosure loss protection
22. We see that we have 388 GB of free capacity with which we can use to create other logical drives or expand existing ones
23. Next assign this logical drive to a host
24. Go to the mappings view (the right hand tab at the top)
25. Right click the LUN and select "additional mapping"
26. Select the desired host from the drop down list and choose add
IBM Blade Center Server Connectivity Module
1. Go to the advanced management console or SCM
2. Move to the configuration page
3. Choose advanced configuration
4. Start a web session directly with the connectivity module
5. Enter the default security information "USERID" and IDK the password
6. First screen deals with port group mapping
7. Port groups allow the blades within our chassis to be segmented into different physical networks
8. The default is all ports in group 1
9. We decide to leave blades 1 thru 10 in group 1 which are associated with uplink ports 1 thru 4
10. Blades 11 thru 14 however need to segmented into a separate network
11. We assign these blades to group 2 and connect them to uplinks 5 and 6
12. Now go to virtual LAN in the left frame
13. Now add VLAN 22 and 23
14. Select group 1 for VLAN 22 and 23
15. Click on the fault tolerance link in the left frame
16. In this case we have a 2nd SCM in bay 2 so we enable switch failover by setting the minimum number of uplinks available which will trigger the failover for each port group
17. Enable group one and choose "1" for "number of links to trigger failover"
18. Enable group two and choose "0"
19. Leave the link aggregation disabled
20. Click apply
21. We will now see how the SCM responds to uplink failures in port group 2
22. The failure of uplink 6 is reflected in the color of the port changing from green to white
23. But no other action is required
24. When uplink 5 fails however, the SCM recognizes that the minimum number of active links has been reached
25. Since the server blades in group 2 no longer have an active uplink available through this module, the SCM disables their internal switch ports
26. This signals the blades' teaming network adaptors to fail over all traffic to the other switch module in the blade center chassis
27. This does not impact traffic in port group one
28. Later when the uplink is restored, the internal ports are re-enabled and traffic returns to normal
IBM Director Boot from SAN Blade Failover Extension
1. In this demo blade 2 is the application server and blade 14 is the spare blade
2. The iSCSI protocol is being used to connect blade 2 to a windows os boot partition on an IBM n series n3700 disk array
3. The blade center management module web interface has a remote control which provides the ability to view and manage a blade server
4. You can define the criteria of hardware conditions which constitutes an actual failure, i.e.., the failure of a CPU, memory DIMM component, or a PFA alert or environmental conditions such as excessive temperature or voltage
5. Upon the receipt of a detected failure on the blade, the alerting capability of the IBM director can generate multiple forms of notification.
6. A popup message appears saying that blade 2 has failed and blade 14 is taking over
7. A red ticker tape will begin to scroll by indicating the same thing
8. You can now see that your remote control session to blade 2 has ended
9. Change remote control to blade 14 and you will see the blade powering up
10. When IBM director detects a critical condition the IBM failover extension will power off the offending blade
11. Next it will detach all san based disks from the failed blade and attach them to a spare blade
12. Finally the spare blade will be powered on
13. Once the spare blade has successfully completed post, the operating system will be loaded along with its installed applications
14. Entire failover takes less than 5 minutes
Setting up a DS3000
1. Takes 30 minutes normally but here's how to do it in 5
2. Launch the storage manager
3. Do an automatic discovery of the storage subsystem
4. Double click on the new un-named storage and you will go to the storage subsystem management window
5. Rename the storage subsystem
6. Setup a storage subsystem password
7. Configure host access
8. If you have any questions there's a link at the top to online help
9. Clicking on the summary tab will provide an overview of your storage subsystems
VMWare,VMotion, HA, and DRS
1. VMotion is the ability to move a running virtual machine from one host to another without taking it down
2. HA makes it so that if one of your virtual machine dies, a virtual machine will be started up on another host
3. DRS allows you to dynamically allocate your virtual machines to your different hosts.
4. If you have a host that's over utilized, drs will automatically move a virtual machine off that host and onto an underutilized host
5. To use VMotion to migrate to a host with less resource utilization, RDP to the host that's over utilized, start Perfmon, start the database with several threads and see if you are running out of resources
6. Right click on the VM and choose migrate, the VMotion wizard pops up, select blade with resources
7. It will validate compatibility, make sure the sans are hooked up, the network is hooked up, it will make sure that the two hosts can talk to each other
8. Select high priority for the task and click next
9. VMotion will start and you'll see a slight drop in performance on the VM
10. VMotion completes in less than 2 minutes with 2 slight dips in performance
11. The RDP connection was never dropped and the database never lost connectivity