For Data ONTAP 6.0 and later, the firmware is downloaded automatically to disks with previous versions of firmware. For information on automatic firmware update downloads, see AUTOMATIC vs. MANUAL FIRMWARE DOWNLOAD.
On configurations using software-based disk ownership both automatic and manual downloads update firmware only for those disks owned by the local node. Disks owned by the partner node are updated when the automatic download or the manual disk_fw_update command is executed on that particular node. Disks that are owned by neither node are not updated. To update all disks, you must assign ownership to the unowned disks before running the disk_fw_update command.
Ignore any warning messages issued while the disk firmware is being updated.
To download the firmware to every disk, use the disk_fw_update command without arguments. To download the firmware to a particular list of disks, specify the disk names in the command. The disk names are in the form of channel_name.disk_ID. For example, if you want to update firmware on disk ID 0, 1 and 3 on adapter 8, enter the following command:
disk_fw_update 8.0 8.1 8.3
The command applies to both SCSI disks and Fibre Channel disks.
If you need to view the current firmware versions, enter the sysconfig -v command. The following example displays a partial output from the sysconfig -v command, where the firmware version for the disk is NA01:
slot 8: Fibre Channel Host Adapter 8 (QLogic 2200 rev. 5, 64-bit, L-port,) Firmware rev: 2.1.20 Host Loop Id: 7 FC Node Name: 2:000:00e08b:00c702 Cacheline size: 8 FC Packet size: 2048 SRAM parity: Yes External GBIC: Yes 0: NETAPP X225_ST336704FC NA01 34.5GB ( 71687368 512B/sect)
- Manual download is a disruptive operation, that makes disks inaccessible for
up to five minutes after the start of its execution, network sessions that
use the node must be closed down before running the disk_fw_update command.
- Is not automatically downloaded to the node's partner node in
an HA pair.
- Is not automatically downloaded to unowned disks on nodes
configured to use software-based disk ownership.
- For Data ONTAP 7.0.1 or later a new registry entry controls how the
automatic firmware download feature works:
If raid.background_disk_fw_update.enable is set to off,
disk_fw_update will run as in previous releases of Data ONTAP.
If raid.background_disk_fw_update.enable is set to on,
disk_fw_update will only automatically update to filesystem disks
contained in RAID4 volumes. Firmware updates for spares and filesystem
disks contained within RAID-DP, mirrored RAID-DP and mirrored RAID4
volumes will be done in a non-disruptive manner in the background after
boot. Firmware download for these disks will be done sequentially by
temporarily taking them offline one at a time for the duration of the download.
Once firmware is updated, the disk will be made online and restored back to
normal operation mode.
During an automatic download to an HA environment, the firmware is not downloaded to an HA partner's disk.
Automatic downloads in HA environments are unsuccessful with certain disk drives. In such cases, you may need to manually execute the disk_fw_update command to update the disks in an HA environment.
When you manually key in the disk_fw_update command, the firmware is:
- Updated on every disk regardless of whether it is on the A-loop,
the B-loop, or in an HA environment.
- If the node is configured in a software-based disk ownership
system, only disks owned by this node are updated.
Data ONTAP 6.1 and later supports redundant path configurations for disks in a non-HA configuration. Firmware is automatically downloaded to disks on the A-loop or B-loop of redundant configurations that are not configured in an HA pair and are not configured to use software-based disk ownership.
When disabled or set to "off", disk_fw_update will update firmware in automated mode just like on previous releases of Data ONTAP. Namely all disks which are downrev will be updated regardless of whether they are SPARE or filesystem disks.
When enabled or set to "on", background disk_fw_update will update firmware in automated mode only to disks which can be offlined successfully from active filesystem raid groups and from the spare pool. For filesystem disks, this capability currently exists within volumes of type RAIDDP, mirrored RAID-DP, and mirrored RAID4. To ensure a faster boot process, no firmware will be downloaded to spares and filesystem disks contained in the above volume types. However, firmware updates for disks within RAID4 volumes will be done at boot. RAID4 volumes can be temporarily (or permanently) upgraded to RAID-DP to automatically enable background firmware update capability.
This provides the highest degree of safety available, without the cost of copying data from each disk in the system twice. Disks are taken offline one at a time and then firmware is updated on them. The disk is made online after the firmware update and a mini/optimized reconstruct happens for any writes which occurred while the disk was offline. Background disk firmware update will not occur for a disk if its containing raid group or the volume is not in a normal state (for example, if the volume/plex is offline or the raid group is degraded). However, due to the continuous polling nature of background disk firmware update, firmware updates will resume once the raid group/plex/volume is restored to a normal mode. Similarly, background disk firmware updates are suspended for the duration of any reconstruction within the system.
If you use the manual disk_fw_update command on a node that belongs to an HA pair, the node downloads the firmware to its disks and its partner's disks, unless the nodes are configured for software-based disk ownership. In that configuration firmware is only downloaded to disks the node owns.
The automatic firmware download only takes place on a node's local disks.
For controller failover configurations, CFO must be enabled and the CFO interconnect must be linked.
Known disk manufacturer limitations on certain disks further contribute to problems with firmware updates. For this reason, Data ONTAP does not allow firmware updates by automatic download to disk models with known limitations when a node is configured in an HA environment.
Disks with known limitations can only accept firmware updates if the disk_fw_update command is executed manually. You may need to enter and run the command yourself. However, no matter which disks your HA configured nodes use, it's safest to update firmware this way.
Use the following procedure to successfully update your disk firmware in an HA environment:
1. Make sure that the nodes are not in takeover or giveback mode.
2. Install the new disk firmware on Node A's disks by issuing the
disk_fw_update command on Node A.
3. Wait until the disk_fw_update command completes on Node A, and
then install the new disk firmware on Node B's disks by issuing
the disk_fw_update command on Node B.
Alternatively, if the registry entry raid.background_disk_fw_update is
enabled then one simply needs to allow the HA partners to update
firmware one disk at a time in automated background mode.
