Isilon Systems IQ S Series Use and care manual

Isilon IQ
S-Series Field Replacement Manual
August 2010
120-0031-01 Rev C

Isilon Systems respects the intellectual property of others.
©2010 Isilon Systems, Inc. All rights reserved.
Isilon, Isilon Systems, OneFS, and SyncIQ are registered trademarks of Isilon Systems, Inc., Isilon IQ, SmartConnect,
SnapshotIQ, TrueScale, Autobalance, FlexProtect, SmartCache, SmartPools, InsightIQ, "HOW BREAKTHROUGHS
BEGIN." and the Isilon logo are trademarks of Isilon. Other product and company names mentioned are the trademarks of
their respective owners.
U.S. Patent Numbers 7,146,524; 7,346,720; 7,386,675; 7,509,448; 7,509,524; 7,551,572; 7,590,652; 7,593,938. China Patent
Number ZL 02816665.5. Other patents pending.
Due to ongoing product development, innovation, and revision, the information contained here is subject to change without
notice. Isilon will publish updates and revisions to this document as needed. The systems documented herein are provided
under End User License Agreements. Please refer to the Agreements for details governing the use of Isilon IQ systems.

Contents
Chapter 1: Field replacement procedures overview ................................................................5
Safety precautions ..............................................................................................................5
General recommendations .................................................................................................5
Contacting Isilon Technical Support ...................................................................................6
Chapter 2: Front panel ................................................................................................................7
Replacing the front panel and hinges .................................................................................7
Chapter 3: Power supply ............................................................................................................9
Replacing a power supply...................................................................................................9
Chapter 4: NVRAM battery assembly ......................................................................................11
Replacing an NVRAM battery assembly...........................................................................11
Chapter 5: Hard drives..............................................................................................................13
Replacing a failed hard drive ............................................................................................13
Adding an SSD to a 5400S node......................................................................................15
Chapter 6: Cooling fans............................................................................................................19
Replacing a fan.................................................................................................................19
Chapter 7: DIMM modules ........................................................................................................23
Replacing a DIMM module ...............................................................................................23
Chapter 8: Expansion cards.....................................................................................................27
Removing the riser card assembly ...................................................................................27
Installing the riser card assembly .....................................................................................29
Replacing the IB/NVRAM Card.........................................................................................31
Replacing the SAS drive controller card ...........................................................................32
Replacing the network interface card ...............................................................................33
Chapter 9: Node chassis...........................................................................................................35
Removing the node top panel...........................................................................................35
Installing the node top panel.............................................................................................35
Chapter 10: SATA cables.........................................................................................................37
Replacing a SATA cable...................................................................................................37

Chapter 11: Air baffle ................................................................................................................39
Removing the air baffle.....................................................................................................39
Installing the air baffle.......................................................................................................40

1
Field replacement procedures overview
Isilon IQ nodes contain a number of components that are easily serviced in the field by trained service personnel.
Note: It is important to replace hardware components as soon as possible after becoming aware of a hardware
failure or impending hardware issue.
Before performing maintenance on Isilon hardware, carefully read and understand all relevant safety information and
recommendations. Contact Isilon Technical Support with any questions about Isilon hardware maintenance.
Safety precautions
Safety warnings describe circumstances under which there is risk of electric shock, bodily injury, fire, damage to Isilon
hardware, or loss of data. Failure to heed these warnings may void the product warranty.
• Ensure that the ambient temperature of the environment in which an Isilon node operates does not exceed the
maximum rated ambient temperature: 35° C (95° F).
• Provide proper airflow to Isilon nodes.
•Ensure that the AC supply circuit for each Isilon node is capable of supplying the total current specified on the node
label.
•Properly ground all AC power supply connections. Pay close attention to connections that are not directly connected
to the branch circuit, such as a node connected to a power strip.
•Before performing maintenance on Isilon nodes, disconnect both power cables.
• Do not overload the branch circuit of the AC power supply providing power to the rack cabinet containing Isilon
nodes. The total rack load must not exceed 80 percent of the branch circuit rating.
• To help protect the system from sudden increases or decreases in electrical power, use a surge suppressor, line
conditioner, or uninterruptible power supply (UPS).
• When installing Isilon nodes in a rack cabinet, always use both front and side rack stabilizers. Installing nodes in a
rack cabinet without stabilizers can cause the rack to tip over and potentially result in personal injury.
• Due to the weight of Isilon nodes, use a minimum of two people to install nodes into a rack cabinet. A 4U Isilon
node with the drives installed can weigh up to 130 lbs (57.97 kg). To minimize the chance of injury during node
installation, 4U Isilon nodes must be installed in a rack cabinet without the hard drives installed in the node.
General recommendations
Before replacing components or performing other maintenance operations on Isilon nodes, review these general
recommendations and best practices.
• Only trained and qualified personnel should install or replace equipment.
• Use only approved replacement parts and equipment.
• Handle all electronic components according to the manufacturer instructions.
5

• Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards and other
static-sensitive system components. You must use a wrist strap attached to the node chassis for grounding when
working on these components.
•Youmayneedtologintothenodeandaccessthesysteminordertoviewstatusandotherinformationaboutdefective
components, or to configure and verify the proper installation of the new components.
•If a drive, power supply, or other component of an Isilon node fails or needs to be replaced, the replacement part kit
must be supplied by Isilon Systems. This replacement policy helps to ensure compatibility between all of the parts.
Inaddition,IsilonSystemsrequeststhatcustomersreturndefectivepartstoIsilonSystemsforanalysisandevaluation.
To return a component to Isilon, contact Isilon Technical Support and follow the instructions provided by the support
engineer. Make sure to include the following details about the failed component on a label or packing sheet:
• Serial number of node
• Technical Support case number (if any)
•Slot number or drive number (if applicable)
• Brief description of the failure circumstances
• Component serial number (if applicable)
Include the label on the package or insert the packing sheet in the package and follow the instructions provided by
the Technical Support Engineer to send the component back to Isilon.
Contacting Isilon Technical Support
Contact Isilon Technical Support if you have any questions about Isilon nodes.
Isilon Technical Support engineers can be contacted using any of the following methods:
Phone: [phone]
Email:
[email protected]
Web: http://www.isilon.com/support
Japan support: 03-5358-7180
Field replacement procedures overview
6©2010 Isilon Systems, Inc.

2
Front panel
The front panel covers the hard drive bays and contains an LCD to display node status information.
General ESD guidelines
Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards, memory, and other
system components. You must use a wrist strap, attached to the node chassis, for grounding.
Required hardware
• Replacement front panel assembly
• ESD grounded wrist strap
Replacing the front panel and hinges
Note: Removing a front panel does not require powering down the node. However, you may need to restart
the LCD on the front panel through the command-line interface when you replace the front panel.
1. Remove the chassis retaining screw that secures the node to the rack cabinet.
2. Slide the node out several inches from the rack cabinet to allow clear access to the node front panel and hinges.
3. Remove the front panel and hinges from the node by removing the four screws that secure the front panel and hinges
to the node chassis.
4. Install the right and left hinge brackets onto the new front panel using the provided shoulder screws.
5. Align the front panel and hinge brackets with the node sidewall standoffs and secure the assembly using the provided
screws.
6. Open the cable latch connector on the inside of the front panel.
7. Connect the flat display cable from node to the cable latch connector.
8. Close the cable latch connector to secure the display cable.
9. Close the node front panel.
10.Secure the node in the rack cabinet.
11.Log in to the node, type isi_lcd_d, and then press ENTER to restart the LCD on the front panel.
7

3
Power supply
Isilon IQ nodes include a dual power supply system to help minimize downtime caused by power source complications. The
two power supply units operate in tandem, each providing half of the amperage draw to the node components.
General ESD guidelines
Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards, memory, and other
system components. You must use a wrist strap, attached to the node chassis, for grounding.
Required hardware
• Replacement power supply
• ESD grounded wrist strap
Replacing a power supply
!
Caution: Never remove both power supply units from a running node.
Note: Replacing a power supply does not require powering down the node. The remaining power unit
provides enough power for the node.
1. Check the LED on each of the power supply units to determine which unit has failed. An amber LED indicates a
failed unit. Disconnect the power cord from the plug of the unit to be replaced.
9

12
1 Power supply LED 2 Power supply release lever
Figure 1. 2U Isilon IQ power supply location
2. Push the power supply release lever to the left to release the unit from the retaining catch.
3. Pull the power supply unit from the power supply bay, using the vertical loop handle.
1
2
1 Vertical loop handle 2 Power supply bay
Figure 2. Removing a power supply from the node
4. Insert a new power supply into the empty power supply bay with the label side facing up.
5. Push the new power supply firmly but carefully into the bay. The retaining latch automatically clicks into place to
secure the power supply.
6. Connect a power cord to the new power supply.
The LED on the power supply turns green when the power supply begins supplying power to the node.
7. Return the old hardware to Isilon Technical Support.
For more information, see
General recommendations on page 5.
Power supply
10 ©2010 Isilon Systems, Inc.

4
NVRAM battery assembly
Isilon IQ nodes have two batteries that provide redundant power to temporarily maintain non-volatile RAM (NVRAM) journal
information in the event of an unexpected power outage and shutdown of the node. The battery units are accessible from the
back of the node.
Failure of an NVRAM battery causes the node to switch to a read-only state as well as generate alerts. Read-only status can
be confirmed on the Node Status page on the web administration interface, on the front panel display, or through the command-
line interface.
After replacing an NVRAM battery, the node performs a quick test to ensure that the new battery is charging properly. It can
take up to seven hours for the battery to fully charge. After charging successfully completes, a full battery test is scheduled
for the next day. A full battery test runs in the background and can take up to 60 hours.
General ESD guidelines
Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards, memory, and other
system components. You must use a wrist strap, attached to the node chassis, for grounding.
Required hardware
• Replacement non-volatile RAM (NVRAM) battery assembly
• ESD grounded wrist strap
Replacing an NVRAM battery assembly
After an alert is issued, it is important to replace the NVRAM battery as soon as possible.
After an alert is issued, it is important to replace the NVRAM battery as soon as possible.
Important: Do not remove both NVRAM batteries at the same time. Removing both NVRAM batteries
puts data at risk in the event of a power loss.
Note: Replacing the NVRAM battery assembly does not require powering down the node. However, verify
that the node is in read-only state prior to replacing the NVRAM batter assembly. Use the Node Status page
on the web interface, front panel LCD display, or isi readonly show from the command-line interface
to verify the node is read-only.
1. On the node back panel, check the LED indicator for each battery assembly to determine which battery has failed.
The LED on the right side of the battery assembly lights red when the battery fails.
11

1
1 NVRAM battery assemblies
Figure 3. NVRAM battery location
2. Lift up the locking tab at the bottom of the battery battery assembly.
2
1
1 Locking tab 2 NVRAM battery
Figure 4. Removing an NVRAM battery assembly
3. Remove the battery assembly out of the node, using the handle. The system software detects the removal of the
battery, and generates alerts.
4. Wait at least 30 seconds for all the alerts to generate.
5. Insert a new battery assembly.
The locking tab clicks into place when the battery assembly is fully inserted.
The battery alert is replaced by an INFO alert stating that the new battery is undergoing a rapid test.
6. Wait five minutes for the rapid battery test to conclude.
If the test concludes successfully, the INFO alert stops, the red Service LED turns off, and the node returns to read-
write mode.
Note: If the battery test does not complete successfully, remove the battery assembly and insert another
battery assembly.
7. Verify that the Charge LED is lit indicating that the new battery is charging.
It may take up to seven hours for the battery to fully charge. After the battery is fully charged, a full battery test is
automatically scheduled for the next day. The full battery test runs in the background and takes 60 hours to complete.
NVRAM battery assembly
12 ©2010 Isilon Systems, Inc.

5
Hard drives
Isilon IQ nodes store cluster data on a series of hard drives. If a hard drive fails, the hard drive must be replaced.
Important: Replacement hard drives must be supplied by Isilon Systems. Hard drives must also have the correct
capacity and type for the node model. If you do not know what kind of hard drive the node requires, contact Isilon
Technical Support.
General ESD guidelines
Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards, memory, and other
system components. You must use a wrist strap, attached to the node chassis, for grounding.
Required hardware
• Replacement hard drive
• ESD grounded wrist strap
Replacing a failed hard drive
Replace failed drives as soon as possible to minimize the potential for data loss.
!
Caution: DonotremoveahealthydrivefromanodeunlessinstructedtoremovethedrivebyIsilonTechnical
Support.
Drive bays are numbered left-to-right, top-to-bottom. For example, the top row of bays on the front of a node are
numbered 1-4.
1. Identify the failed hard drive.
If you already know which hard drive to replace, proceed to the next step. To determine which hard drive to replace,
complete the following substeps.
a. Log in to the node.
b. Run the isi devices -a status -d <node ID number> command.
The system displays text similar to the following example output:
Node 4, [ OK ]
Bay 1 Lnum 8 [HEALTHY] SN:A603SPEE /dev/da0
Bay 2 Lnum 7 [HEALTHY] SN:A6082A2E /dev/da1
Bay 3 Lnum 6 [HEALTHY] SN:Y60PFJWE /dev/da2
Bay 4 Lnum 5 [HEALTHY] SN:A6086AVE /dev/da3
Bay 5 Lnum 4 [HEALTHY] SN:Y604H2NE /dev/da4
Bay 6 Lnum 3 [HEALTHY] SN:Y604XD5E /dev/da5
Bay 7 Lnum 9 [HEALTHY] SN:Y604HBGE /dev/da6
Bay 8 Lnum 10 [HEALTHY] SN:Y603E6GE /dev/da7
13

Bay 9 Lnum 0 [REPLACE] SN:Y60PFPEE /dev/da8
Bay 10 Lnum 11 [HEALTHY] SN:A60877NE /dev/da9
Bay 11 Lnum 11 [HEALTHY] SN:A60877NE /dev/da10
Bay 12 Lnum 11 [HEALTHY] SN:A60877NE /dev/da11
In this example output, the hard drive in Bay 9 shows a REPLACE state, indicating that it needs to be replaced.
2. Remove the chassis retaining screw that secures the node to the rack cabinet.
3. Slide the node out several inches from the rack cabinet to allow the front panel to open freely.
4. Open the node front panel.
5. Locate the failed hard drive, release the drive locking handle, and remove the failed drive.
2
3
1
1 Locking handle 3 Hard drive bay
2 Hard drive
Figure 5. Removing a hard drive
6. With the locking handle open, insert a drive into an empty drive bay by sliding the drive along the bay rails until it
stops.
!
Caution: Do not force the drive into the drive bay. Forcing the drive into the drive bay could result in
irreparable damage to both the drive and the drive bay.
Note: SSDs can only go in Bay 1 (the upper-left drive bay). Non-SSD drives can go in any drive bay.
You can identify an SSD by the color of the drive tray locking handle; SSDs have silver handles, whereas
all other drive types have black handles. SSDs are also smaller 2 1/2" drives.
7. Gently push the locking handle on the drive until the drive clicks into place.
8. Close the node front panel.
9. Slide the node back into the rack cabinet.
10.Secure the node to the rack cabinet using the chassis mounting screw.
11.On the command-line interface, run the isi devices -a add -d <node ID number>:<drive bay
number> command.
12.Complete one of the following options based on the result of the previous step.
• If the inserted drive has been properly formatted, the following message appears:
The add operation succeeded. A OneFS-formatted drive was found in
bay
<drive bay number> and is being attached to the file system. Wait a few
minutes and re-run 'isi devices' to verify that the attach
operation completed successfully.
Wait a few minutes, and then run the isi devices command again to verify that the drive was added
successfully.
•If the inserted drive has encountered an add-operation failure, an error message appears. If the message indicates
that the failure is the result of an unrecognized drive or a previously used drive, or that the removed drive has
Hard drives
14 ©2010 Isilon Systems, Inc.

been re-inserted, run the isi devices -a format -d <node ID number>:<drive bay number>
command to format the drive.
• If the inserted drive has not been previously formatted, the following message appears:
This drive does not appear to be pre-formatted with OneFS.
Formatting this drive will destroy any existing data and attach
it to OneFS. Do you wish to continue formatting this drive?
To format the drive, type yes. To cancel the formatting process, type no.
• If the inserted drive has been used in another node, the following message appears:
This drive appears to contain OneFS data not recognized by this
node. Formatting this drive will erase the existing data,
completely re-format the drive, and attach it to OneFS. Do you
wish to continue formatting this drive?
To format the drive, type yes. To cancel the formatting process, type no.
• If the inserted drive is the same drive that was removed, the following message appears:
This drive has been removed from the cluster. It is
likely that the drive is faulty and the suggested action is to
replace the drive. Re-formatting the drive will erase all data
and force it back into service. Are you sure you wish to continue
formatting this drive?
To format the drive, type yes. To cancel the formatting process, type no.
Note: If the add-operation failure is not the result of an unrecognized drive or a previously used drive,
or if the removed drive has been re-inserted, follow the instructions provided in the error message or
contact Isilon Technical Support.
13.Return the old hardware to Isilon Technical Support.
For more information, see
General recommendations on page 5.
Adding an SSD to a 5400S node
This procedure describes how to upgrade an Isilon IQ 5400S node to a fully-supported Isilon IQ 5000S-SSD node.
Prerequisite:
Note: This upgrade is available only for existing customers with S-Series nodes already deployed in the
field.
Current S-Series customers can upgrade to add solid state drives (SSDs) to existing nodes. Customers can purchase the
upgrade option via an upgrade kit, which is a single line item on a purchase order. This order initiates Technical Support
and Professional Services to contact the customer and plan the upgrade.
TheTechnicalSupport team guides the customerthroughthepreliminaryphase of the upgrade, whichrequiresupgrading
the OneFS operating-system software, smartfailing the disk drives in Bay 1 of each node on the cluster, and running an
Isilon FlexProtect operation. After these steps are completed, the customer must schedule an appointment with a field
representative from Isilon Technical Support or Isilon Partner Services, who will perform the remaining upgrade steps
at the customer site.
Preliminary tasks to be completed by the customer
Instruct the customer to complete the following steps before the field representative arrives at the customer site.
Note: SSDs can only go into Bay 1 of a 2U node.
1. Upgrade the cluster to OneFS 5.5.4.21.
Hard drives
©2010 Isilon Systems, Inc. 15

2. Smartfail the disk drives that will be replaced by running the following command:
isi_for_array -S isi devices -f -a smartfail -d :bay1
The system smartfails the disk drive that is located in Bay 1 of each node in the cluster.
3. Put the Bay 1 disk drives in a REPLACE state by running the following command:
isi restripe start flexprotect
Note: You can speed up the smartfail process--at the risk of degraded performance--by running the
preceding command with the high option:
isi restripe start flexprotect high
4. View information about the Bay 1 drive of each node by running the following command:
isi_for_array -S isi devices | grep "Bay 1 "
The system displays text similar to the following example output:
Node-2: Bay 1 Lnum 11 [REPLACE] SN:Y446D4EE /dev/twed0
Node-1: Bay 1 Lnum 0 [REPLACE] SN:KRVN65ZA /dev/da12
...
5. In the system output from the preceding step, verify that each of the Bay 1 disk drives is in a REPLACE state.
Upgrade procedure to be completed by the field representative
After the customer has completed the preliminary tasks, the field representative arrives at the customer site with the new
hardware and product labels to perform the following procedure.
NoteV: Steps 1-10 must be repeated for each node on the cluster.
1. Update the system configuration number by running the following command:
echo <NPEFM@OVNCFS> > /etc/isilon_system_config; isi_dongle_sync –p
2. Replace the SAS drive in Bay 1 with an SSD.
3. Restart the isi_drive_d daemon by running the following command:
isi_drive_d restart
4. Add the drive by running the following command:
isi devices -f -a add -d :bay1
5. Run the following command:
sysctl efs.bam.layout.ssd. | grep mirror
The system displays text similar to the following example output:
efs.bam.layout.ssd.inode.mirrors: 1
efs.bam.layout.ssd.sys_btree.mirrors: 1
efs.bam.layout.ssd.dfm_btree.mirrors: 1
efs.bam.layout.ssd.ifm_btree.mirrors: 1
efs.bam.layout.ssd.ifm_ext.mirrors: 1
efs.bam.layout.ssd.qab.mirrors: 1
efs.bam.layout.ssd.stf_delta.mirrors: 1
6. In the system output from the preceding step, verify that each of the efs.bam.layout.ssd sysctls is set to the default
value of 1.
7. Run the following command:
sysctl -n efs.drv.state
The system displays text similar to the following example output:
{(0, da12s1e, 5eabfa6f5018e2a3 4b58f26e0000fe2e, UP, softfail=0, ssd=0),
(1, da11s1e, 3fef9cd1454c0c51 4b58f26e0000fe3e, UP, softfail=0, ssd=0),
(2, da10s1e, 2744b4982f006b74 4b58f26e0000fe41, UP, softfail=0, ssd=0),
Hard drives
16 ©2010 Isilon Systems, Inc.
In Step 1, replace [model_number] with the model number of the device being updated.

(3, da9s1e, 7a5802340b408637 4b58f26e0000fe45, UP, softfail=0, ssd=0),
(4, da8s1e, 46a0fc0443b7f537 4b58f26e0000fe49, UP, softfail=0, ssd=0),
(5, da7s1e, 20a7f22616174ba3 4b58f26e0000fe4d, UP, softfail=0, ssd=0),
(6, da6s1e, 3949624b7b8a2d9d 4b58f26e0000fe51, UP, softfail=0, ssd=0),
(7, da5s1e, 27a1f7ac4657dfda 4b58f26e0000fe56, UP, softfail=0, ssd=0),
(8, da4s1e, 59198b8b1237d60f 4b58f26e0000fe5a, UP, softfail=0, ssd=0),
(9, da3s1e, 3ea11ef23fc52232 4b58f26e0000fe60, UP, softfail=0, ssd=0),
(10, da2s1e, 25bebd5106739b19 4b58f26e0000fe66, UP, softfail=0, ssd=0),
(11, da1s1e, 213df93176467a83 4b591956000cf778, UP, softfail=0, ssd=1)}
8. In the system output from the preceding step, verify that the ssd value for the drive is set to 1.
9. On the front panel, replace the existing 5400S model nameplate with the new 5000S-SSD nameplate.
10.Attach the new 820-0001-01 SKU label over the existing SKU label.
Note: Be sure to cover the lower SKU label, as shown in the following figure. Do not cover the upper
serial number label.
2
1
1 Existing SKU label 2 Replacement SKU label
Figure 6. Replacing the SKU label
11.Repeat steps 1-10 until all nodes on the cluster have been updated with SSDs. When all of the nodes have been
updated, continue to the next step.
12.From any node on the cluster, perform the following actions.
These actions will be applied to all nodes on the cluster:
1. Run the following commands:
sysctl efs.bam.layout.min_free_drive_blocks=3000
echo efs.bam.layout.min_free_drive_blocks=3000 >> /etc/mcp/override/
sysctl.conf
Running these commands prevents a known issue in which ENOSPC errors are incorrectly reported if SSDs fill
with metadata while space is available on other drives. For more information, see Isilon knowledgebase article
201026.
2. Updatethe image of the node that appearsinthe web administration interface by running thefollowingcommand:
isi_for_array 'cp /usr/local/www/ui/images/hwIQ5400S.jpg /usr/local/www/
ui/images/hwIQ5000s-ssd.jpg'
3. Restripe the metadata to the SSDs by running the following command:
isi restripe start autobalance high
13.Record all of the node serial numbers and new system configuration numbers in the Salesforce case.
14.Pack the SAS drives that were removed from the nodes into the SSD drive boxes, apply the shipping labels, and ship
them to Isilon.
For more information, see General recommendations on page 5.
Hard drives
©2010 Isilon Systems, Inc. 17

6
Cooling fans
Isilon IQ nodes contain three cooling fans to prevent the node components from overheating.
General ESD guidelines
Always follow electrostatic discharge (ESD) prevention procedures when removing and replacing cards, memory, and other
system components. You must use a wrist strap, attached to the node chassis, for grounding.
Required hardware
• Replacement cooling fan
• ESD grounded wrist strap
•#2 Phillips screwdriver or #2 JIS driver
Replacing a fan
Note: Replacing a fan does not require powering down the node.
1. Remove the top panel from the node.
For more information, see Removing the node top panel on page 35.
Note: Removing the top panel from a node temporarily places the node into a read-only state. The node
returns to a read/write state when the top panel is replaced.
2. Identify the fan you want to replace. A red Fan Fault LED on the fan control board identifies the failed fan.
19

1
1 Cooling fans
Figure 7. 2U Isilon IQ fan location
3. Unlatch the green release lever on the fan assembly and pull the fan straight up.
!
Caution: Keep fingers clear of fan blades while removing the fan.
2
1
3
1 Fan latches 3 Fan latch slot
2 Release lever
Figure 8. Removing a fan
4. Slide the new fan into the fan sleeve in the correct orientation.
5. Lift the release lever to engage the fan latch on the release lever side of the fan.
6. Press the fan straight down and secure the fan latch on the opposite side of the release lever.
The fan should begin spinning and the Fan Fault LED should turn off.
7. Replace the node top panel.
For more information, see Installing the node top panel on page 35.
Cooling fans
20 ©2010 Isilon Systems, Inc.