Scale Up X200 - Add Drives

Important Information - ActiveScale OS Requirements

ActiveScale X200 Scale Up requires ActiveScale OS 6.3 and above. Upgrade X200 systems before following the scale up procedures below.

Disk Installation Configuration

TwinServer/JBOD Bottom Top
Slots 0 - 14 (occupied) 45 - 59 (occupied)
Slots 15 - 28 (empty) 60 - 74 (empty)
Slots 29 - 44 (empty) 75 - 89 (empty)
TwinServer/JBOD Bottom Top
Slots 0 - 14 (occupied) 45 - 59 (occupied)
Slots 15 - 28 (occupied) 60 - 74 (occupied)
Slots 29 - 44 (empty) 75 - 89 (empty)
TwinServer/JBOD Bottom Top
Slots 0 - 14 (occupied) 45 - 59 (occupied)
Slots 15 - 28 (occupied) 60 - 74 (occupied)
Slots 29 - 44 (occupied) 75 - 89 (occupied)

Disk Installation

Adding Disks to 4U90 Twin Server

Add sets of 15 disks to the X200 server storage compartment.

Server Installation Order
Bottom Server

  • 15 disks for each range of Slots: 0-14, 15-29 and 30-44.

  • Disks should be added per 15 in diskgroup A, B, and C order (see Figure 1).
Top Server

  • 15 disks for each range of Slots 45-59, 60-74 and 75-89.

  • Disks should be added per 15 in diskgroup A, B, and C order (see Figure 2).

Figure 1: Bottom Server - Diskgroup A,B,C

Figure 2: Top Server - Diskgroup A,B,C

Follow the Disk Installation Procedure below to add disks to the X200 Twin Server.

Adding Disks to a JBOD

Caution:  

Server Installation Order
Bottom Server

  • 15 disks for each range of Slots: 0-14, 15-29 and 30-44.

  • Disks should be added per 15 in diskgroup D, E, and F order (see Figure 3).
Top Server

  • 15 disks for each range of Slots 45-59, 60-74 and 75-89.

  • Disks should be added per 15 in diskgroup D, E, and F order (see Figure 4).

Figure 3: Bottom Server - Diskgroup A,B,C

Figure 4: Top Server - Diskgroup A,B,C

Follow the Disk Installation Procedure below to add disks to a X200 JBOD.

 

Caution: Use extreme caution when removing or replacing disk drives near the expansion modules. There is very little distance between the drive and the expansion module and damage may occur if care isn't taken.

Caution: Except for short periods of time, such as swapping hard drives, do not operate the server with the hard drive bays empty. The carrier design promotes proper air flow through the chassis.

Tools Required: None

  1. Pull the storage enclosure forward out of the chassis.

  2. Slide the release button on the drive carrier, which opens the carrier handle.

  3. Use the drive carrier handle to pull the drive up out of the chassis.

  4. Under the main the carrier handle, find and then lift the breakout lever and pull out the side of the carrier.

  5. Remove the dummy drive from the carrier.

  6. Insert the hard drive into the drive carrier. Orient the drive by matching the two alignment pins on the side and by noting the notch in the carrier for the HDD connectors. Close the side of the carrier until it snaps in place.

  7. Slide the carrier assembly into its spot in the chassis until it clicks into place. Then press the release button and push the drive carrier handle down into the locked position.

Provision the Newly Added Disks

Important Information

The system must be in a stable state before scaling up. The scale up job checks data safety levels when it starts. If the system is freshly installed, or if you just completed an initial scale up, wait 10-15 minutes for the system to stabilize and start recording data safety levels. Systems that have been running for some time may have a data safety issue as reported in the ActiveScale System Management Interface. You must resolve any data safety issues before attempting to scale up. If the data safety of the system is not optimal or there are failed drives, fix the issues by replacing the drives before starting this procedure.

WARNING:  

Once you have installed the new disks in groups of 15, do the following:

  1. Verify the health of the X200:

    1. Make sure all nodes are online.

    2. Click Configuration > System.

    3. In the General Info section, under Health Check, click Run.

    4. Monitor the status of the health check job in the Jobs pane.

    5. If the health check job has errors, stop. Contact Quantum Support for help.

  2. Scale up is run per rack. To determine which racks the current system has, log into one of the nodes on the X200 and run following line in sherpa. The names of the racks will be printed:

    print "\n".join([rack['name'] for rack in cli.marvin.location.list(query='locationtype=RACK')])

    Now you will have to determine in which racks the drive expansions, or JBOD attachments, were done. For each of those racks, you will have to run following code snippet. Please wait for the code to have finished successfully before running it for the next rack! For each run of the snippet you’ll first have to change the value of the rack_name variable to the name of the rack that you want to scale up now. Put the rack name in (single or double) quotes.

     

    Copy 
    %cpaste
    rack_name = "<rack_name>"
    import json
    from marvin_lib.network import ha
    from marvin_lib.task import wait_for_tasks
    env_guid = cli.marvin.environment.list()[0]['guid']
    rack_location = cli.marvin.location.guid(rack_name)
    machines_in_rack = cli.marvin.machine.list(query='location=%s' % rack_location)
    machine_guid = machines_in_rack[0]['guid']
    task = json.loads(ha.post('/model/environment/{env_guid}/action/add_storage_capacity/?local_machine_guid={machine_guid}'.format(env_guid=env_guid, machine_guid=machine_guid)))['task']
    print wait_for_tasks([task], timeout=3600, terminate_tasks=False, stop_on_error=False)
    --