Highly available storage for the Origin200 FailSafe network server (OFNS) system is provided by mirrored SCSI disk drive modules installed in the system's Origin Vaults. This mass storage is supported by two differential PCI SCSI controllers that install in each Origin200 host (four total). In addition, each host uses one PCI Ethernet controller board for public Ethernet access.
Two mirrored disk modules come standard in each of the two Origin Vaults (four total). Each Origin200 host has one system disk installed. Optional additional disks may be ordered with your system or installed at a later date.
You have to remove the Origin200 server before you can remove and replace a faulty PCI board or perform certain other types of service. Instructions for properly removing an enclosure from the OFNS rack are covered in “Removing an Enclosure From the OFNS Rack”.
Read the following information before you add, remove, or replace a disk drive. You should not need to remove the enclosure to replace a drive.
The SCSI device IDs of the 3.5-inch bays in the Origin200 and Origin Vault are numbered 1 through 6, counting from the right to the left. Figure 3-1 shows a removal and replacement example for a drive in position 6.
Drives 6 and 5 in the Origin Vaults (closest to the power button) are always the mirrored high-availability disks and should never be moved to a different location within the enclosure. See “Mirrored Disk Replacement Example” for more information. Only the Origin Vaults can hold the mirrored highly available disks.
Data disks ordered from Silicon Graphics are preconfigured; you can install them in bays 1 through 5 in the Origin200 with no additional configuration.
There are very specific rules for use of third-party disks in the Origin200 and Origin Vault. See the Origin200 and Origin GIGAchannel Owner's Guide (p/n 007-3708-xxx) for information before installing any third-party drives.
 | Note: Save any blanking plates you remove. If you ever remove a peripheral and do not replace it with another one, you should reinstall the blanking plate to ensure proper cooling, air flow, and electromagnetic interference (EMI) shielding in the enclosure. |
To add or replace a disk in either an Origin200 or Origin Vault enclosure you must shut it off. You should always minimize the time this takes, because shutting it off compromises the high availability of the rack system.
As the superuser (root), use the command
# init 0
to properly shut down the Origin200 before disk installation or replacement.
See the section “Restoring a Failed OFNS System Disk” for useful information on replacing a failed system disk in an Origin200.
In the Origin Vaults you can use the power button on the front to power off the enclosure. See “Adding or Replacing a Drive in the Origin Vault”.
Note that if you want to replace one of the mirrored disks you should read the information in the section “Mirrored Disk Replacement Example”. If you are adding a disk that is not part of a mirrored volume, you must still follow the guidelines for restoring any outdated data that affects the mirrored disks in the vault. See “Updating vol1 in Origin Vault 1” for information and an example.
General information on a system's disks, volume sizes, or XLV configuration is covered in the IRIX Admin: Disks and Filesystems manual (p/n 007-2825-xxx).
The IRIS FailSafe Administrator's Guide (p/n 007-3109-xxx) has other information you may need if you wish to modify or expand your system's disks.
Before replacing or adding a disk to the Origin Vault enclosure, you must first turn the power off to the enclosure. You can shut down the power to the Origin Vault using the front power button. Note that any time you do this (for whatever reason) you should turn power back on as soon as possible. See “Powering the System On and Off” in Chapter 1 if you need more information.
As you will note in Figure 3-2, SCSI controller 2 in each Origin200 connects to Origin Vault 1 and each SCSI controller 3 connects to Origin Vault 2.
Mirrored disks always go in the Origin Vaults used with your OFNS system rack. In the following example we assume that a failed disk in Origin Vault 1 must be replaced.
Assume that the failed disk is drive ID 6 in Origin Vault 1 (dks2d6s7). This drive is in vol2 (plex 0). If a step uses exact syntax, the syntax has been modified to fix this specific disk failure example.
Delete the plex (or volume element) containing the broken disk from the volume (in this case vol2).
# xlv_mgr
xlv_mgr> detach plex vol2.data.0 badplex
If the deletion is successful, go on to “Physically Replacing the Drive in vol2” and continue the procedure.
If the failed disk is unresponsive and causes the detachment attempt to fail, input the following to xlv_mgr:
xlv_mgr> detach -force plex vol2.data.0 badplex
xlv_mgr> delete object badplex
If the detachment operation still fails, go on to step 2.
Forcibly unmount the filesystem
# umount -k /fs2
Save the volume configuration using xlv_mgr.
Delete the volume object:
xlv_mgr> delete object vol2
Remake the volume without the broken disk.
Identify the enclosure with the failed drive (Origin Vault 1 in this example).
Turn off power to Origin Vault 1. See “Powering the System On and Off” in Chapter 1 if you need more information.
Wait 10 seconds. This wait time is important, as it ensures the failed drive doesn't receive additional damage.
Physically remove the failed disk drive and install the replacement disk. See Figure 3-1 for an example.
Power the Origin Vault 1 back on.
If I/O writes occur to vol1 in Origin Vault 2 during the time that Origin Vault 1 is powered off, then vol1 will need to be updated/repaired. Use xlv_mgr to determine if part of vol1 is outdated or [offline] by entering the following command:
xlv_mgr> show kernel
If the output shows [offline] then disk ID5 in Origin Vault 1 contains outdated data. If part of vol1 is [offline], use xlv_mgr to put the affected volume element back on line. In our example, drive dks2d5s7 would have been [offline] due to the power outage. This drive is plex 0 of volume 1. In this example, you would enter the command:
xlv_mgr> change online vol1.data.0.0
Use the following information to provide the replacement drive with all the proper XLV volume elements.
Use fx to partition the new drive exactly as the old drive.
Note: The new drive must be repartitioned the same as the failed drive. Create a plex (volume element) on the new disk drive.
# xlv_make
xlv_make> plex newplexname
xlv_make> ve dks2d6s7Attach the plex (or attach/insert the ve) back to the volume.
# xlv_mgr
xlv_mgr> attach plex newplexname vol2.data
Use the instructions in this section to remove an Origin200 server before replacing a PCI board or performing other service. These instructions can be used to remove any of the enclosures in the OFNS rack system. This should be necessary only when service or replacement of an internal component is required. See “Replacing a PCI Board in an Origin200 Host” for an example. Replacement or installation of a disk does not require removal of the enclosure from the rack. See “Installing or Replacing a Disk Drive” for an example.
 | Caution: To maintain stability, remove only one enclosure at a time. Before removing an enclosure, ensure that all other enclosures in the rack are properly secured with their mounting hardware. |
Use the following instructions to shut down the enclosure, detach its cables, and remove it from the OFNS rack:
Turn off the power to the enclosure.
Press the front power switch to shut off power to the drive bays, PCI backplane, main system board and CPU daughter card (see Figure 3-3).
Turn off the master power switch and unplug the power cable.
Open the rear door of the rack and disconnect the appropriate SCSI cables, public and private Ethernet cables, dumb terminal cable (if attached to serial port 1), the reset cable attached to serial port 2, and the reset cable attached to the eight-pin module system controller port (see Figure 3-4).
Using your fingers or a Phillips head screwdriver, undo the two captive screws that hold the enclosure to the rear retaining bracket (see Figure 3-5).
Go to the front of the system and pull the plastic side caps off both sides of the enclosure.
Undo the four screws that hold the enclosure to the front rails.
Slide the “drawer” and enclosure gently out of the cabinet (about 14 inches [35.5 cm]) until the drawer stops.
With one person on either side of the enclosure, gently lift it up and forward off its foam drawer pad and place it on a clean work surface, (see Figure 3-6).
Go on to the section “Replacing a PCI Board in an Origin200 Host” or see the Origin200 and Origin200 GIGAchannel Maintenance Guide if you need to do other procedures.
Each of the Origin200 host servers in the OFNS rack system comes with PCI boards installed. If a PCI board malfunctions or must be replaced for any reason, use the information in the Origin200 and Origin200 GIGAchannel Maintenance Guide (p/n 007-3709-xxx) to properly complete the procedure.
Prepare the system for servicing by removing it from the rack using the information in “Removing an Enclosure From the OFNS Rack”. Always use two people to lift and carry the enclosure.
You will need a #2 Phillips screwdriver to remove and replace a PCI board; an electrostatic discharge (ESD) wrist strap is provided.