Professional Documents
Culture Documents
Netapp Cluster For NDMP
Netapp Cluster For NDMP
If you have an existing two-node cluster that uses cluster network switches, you can apply the switchless-cluster networking
option and replace the switches with direct, back-to-back connections between the nodes. This is a non-disruptive operation.
The procedure you use depends on whether you have two dedicated cluster-network ports on each controller (as required on
most systems) or a single cluster port on each controller (a supported option on FAS22xx storage systems).
Choices
• Transitioning to a two-node switchless cluster (systems with redundant cluster-network connections) on page 1
• Transitioning to a two-node switchless cluster (FAS22xx systems with a single cluster-network connection) on page 11
• Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) on page 14
• The cluster must be healthy and consist of two nodes connected by cluster switches, and the nodes must be running Data
ONTAP 8.2 or later.
The switchless cluster feature cannot be used with more than two nodes.
• Each storage system must have two dedicated cluster ports providing redundant cluster-network connections.
This is the supported configuration for most storage systems. FAS22xx systems support an optional configuration using a
single cluster-network connection due to the limited number of ports.
• Command output examples in this section are from Data ONTAP 8.2.x, which differ from the command output in Data
ONTAP 8.3 and later.
See Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) on page 14
for example command output for Data ONTAP 8.3 and later.
215-07834_D0 November 2014 Copyright © 2014 NetApp, Inc. All rights reserved. 1
Web: www.netapp.com • Feedback: doccomments@netapp.com
Steps
Example
This example and the others in this procedure show systems that are using e0a and e0b as cluster ports. Your system might
be using different cluster ports due to its particular requirements.
2. Disable the –auto-revert parameter on each of the cluster LIFs on the cluster ports on both nodes:
network interface modify -vserver vserver-name -lif lif_name -auto-revert false
This enables you to then move the LIFs temporarily so that you can remove the cable connections from the switches.
Example
The following illustration shows a simplified view of the two nodes, node1 and node2:
• Each node uses ports e0a and e0b for the cluster connections.
• Each port has a cluster LIF (one with the defined role of cluster) associated with it.
• Each node is configured with a Storage Virtual Machine (SVM, formerly known as Vserver), node1 and node2.
• For the sake of simplicity, only one cluster switch is shown, ClusterSwitch1.
3. Select one of the cluster ports on the first node, and then migrate the interfaces from that port to the other cluster port on the
same node:
network interface migrate –vserver vserver-name -lif lif_name –dest-node destination-node-
name –dest-port port-name
Example
The following illustration and command shows the migration of cluster LIFs from port e0a on node1 to port e0b:
network interface migrate –vserver node1 -lif clus1 –dest-node node1 –dest-port e0b
5. On the second node, select the corresponding cluster port, and then migrate interfaces from that port to another cluster port
on the same node:
network interface migrate –vserver vserver-name -lif lif_name –dest-node destination-node-
name –dest-port port_name
Example
The following command and illustration shows the migration of LIFs from port e0a on Node2 to port e0b:
network interface migrate –vserver node2 -lif clus1 –dest-node node2 –dest-port e0b
7. Disconnect the cables that link the switches from the ports vacated in steps 3 on page 3 and 5 on page 4.
Example
In the example, the cables are disconnected from port e0a on each node and cluster traffic continues through the switch and
port e0b on each node:
8. Cable the ports that were formerly connected to the switch to each other, creating a direct connection between the
controllers.
Example
In the following example, e0a on node1 is connected to e0a on Node2:
9. Revert the LIFs that were migrated in steps 3 on page 3 and 5 on page 4 back to their home ports:
network interface revert –vserver vserver-name -lif LIF-name
Example
The following example shows the command and illustrates the LIF revert:
network interface revert –vserver node1 -lif clus1
network interface revert –vserver node2 -lif clus1
Example
11. Change to the advanced privilege level from either node's prompt:
set -privilege advanced
You can respond y when prompted to continue into advanced mode and get to the advanced mode prompt *>.
14. Select one of the remaining ports connected to the switch on the first node, and then migrate the interfaces from that port to a
port that uses the direct connection:
network interface migrate –vserver vserver_name -lif lif_name –dest-node destination-node-
name –dest-port port_name
Example
16. Select the corresponding cluster port on the second node, and then migrate interfaces from that port to the port that usesthe
direct connection:
network interface migrate –vserver vserver_name -lif lif_name –dest-node destination-node-
name –dest-port port_name
Example
The following example shows the command and illustrates the LIF migration from port e0b on Node2 to port e0a:
network interface migrate –vserver node2 -lif clus2 –dest-node node2 –dest-port e0a
Example
18. Disconnect the cables that link the switches to the ports that no longer have traffic.
Example
In the following example, the cables to the switch are removed and cluster traffic continues through the direct connection
between the e0a ports:
19. Cable the ports that were formerly connected to the switch to each other, creating a direct connection between the
controllers.
Example
Traffic on these ports can now travel through the back-to-back connection.
Example
The following example shows the command and illustrates the LIF migration:
network interface revert –vserver node1 -lif clus2
network interface revert –vserver node2 -lif clus2
Example
22. Reenable the –auto-revert parameter on each of the cluster LIFs on the cluster ports on both nodes:
network interface modify -vserver vserver_name -lif lif_name -auto-revert true
23. Verify that the cluster port LIFs are operating correctly:
network interface show
24. Check the cluster status of the nodes from the system console of either node:
cluster show
25. Verify cluster connectivity with the cluster ping-cluster local command:
You can respond y when prompted to continue into advanced mode and get to the advanced mode prompt *>.
The command output should show connectivity between all cluster ports.
• The cluster must be healthy and consist of two nodes connected by a cluster switch, and the nodes must be running Data
ONTAP 8.2 or later.
The switchless cluster feature cannot be used with more than two nodes.
• Each storage system must use a single, dedicated 10-GbE cluster port that provides the cluster-network connection.
This is a supported configuration option for FAS22xx systems only.
• Command output examples in this section are from Data ONTAP 8.2.x, which differ from the command output in Data
ONTAP 8.3 and later.
See Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) on page 14
for example command output for Data ONTAP 8.3 and later.
Steps
1. Check the cluster status of the nodes from the system console of either node:
cluster show
2. Check the status of the HA pair from the system console of either node:
Transitioning to a two-node switchless cluster (FAS22xx systems with a single cluster-network connection) 11
storage failover show
Example
The output should be similar to the following:
You can respond y when prompted to continue into advanced mode and get to the advanced mode prompt (*>).
It does not matter which node is the target node. When it is taken over, the target node automatically reboots and displays
the Waiting for giveback... message.
The active node is now serving data for the partner (target) node that has been taken over.
7. Wait at least two minutes after takeover of the impaired node to ensure that the takeover was completed successfully.
8. With the target node showing the Waiting for giveback... message, shut it down.
The method you use depends on whether remote management through a Service Processor (SP) is used.
9. On each controller, disconnect the cable that connects the 10-GbE cluster port to the cluster switch.
10. Connect the cluster port of each controller to the cluster port on the other controller, so that the controllers have a back-to-
back network connection.
The target node boots and displays the Waiting for giveback... message.
13. Monitor the progress of the giveback operation by using the storage failover show-giveback command.
14. After the giveback operation is complete, confirm that the HA pair is healthy and takeover is possible by using the storage
failover show command.
Example
The output should be similar to the following:
15. Verify that the cluster port LIFs are operating correctly:
network interface show -role cluster
Example
16. Check the cluster status of the nodes from the system console of either node:
cluster show
17. Verify cluster connectivity by using the cluster ping-cluster local command:
Transitioning to a two-node switchless cluster (FAS22xx systems with a single cluster-network connection) 13
You can respond y when prompted to continue into advanced mode and get to the advanced mode prompt (*>).
The command output should show connectivity between all cluster ports.
• The cluster must be healthy and consist of two nodes connected by cluster switches, and the nodes must be running Data
ONTAP 8.3 or later.
The switchless cluster feature cannot be used with more than two nodes.
• Each storage system must have four dedicated cluster ports, per controller, providing redundant cluster-network connections.
This is the supported configuration for most storage systems that support four cluster interconnects.
Steps
1. Check the status of the nodes in the cluster to make sure health and eligibility are true:
Example
The command produces output similar to the following:
If... Then...
All nodes show true for both health Proceed to the next step.
and eligibility.
Any node shows false for health or Correct the problem and proceed to the next step.
eligibility.
Example
This example and the others in this procedure show systems that are using e0a, e0b, e0c, and e0d as cluster ports. Your
system might be using different cluster ports due to its particular requirements.
4. Disable the –auto-revert parameter on each of the cluster LIFs on the cluster ports on both nodes:
network interface modify -vserver vserver_name -lif lif_name -auto-revert false
This enables you to then move the LIFs temporarily so that you can remove the cable connections from the switches.
Example
The following illustration shows a simplified view of the two nodes, node1 and node2:
• Each node is using ports e0a, e0b, e0c, and e0d for the cluster connections.
• Each port has a cluster LIF (one with the defined role of cluster) associated with it.
• Each node is configured with a Storage Virtual Machine (SVM, formerly known as Vserver), node1 and node2.
• The following example shows the connections to both ClusterSwitch1 and ClusterSwitch2.
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 15
The following steps in this example disable auto-revert on the LIFs:
5. Select one of the cluster ports on the first node and migrate the interfaces from that port to the other cluster port on the same
node:
network interface migrate –vserver vserver-name -lif lif-name –dest-node destination-node-
name –dest-port port-name
Example
The following illustration and command shows the migration of cluster LIFs from port e0a on node1 to port e0c:
network interface migrate –vserver Cluster -lif node1_clus1 –dest-node node1 –dest-port e0c
7. On the second node, select the corresponding cluster port and migrate interfaces from that port to another cluster port on the
same node:
network interface migrate –vserver vserver-name -lif lif-name –dest-node destination-node-
name –dest-port port-name
Example
The following command and illustration shows the migration of LIFs from port e0a on node2 to port e0c:
network interface migrate –vserver Cluster -lif node2_clus1 –dest-node node2 –dest-port e0c
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 17
8. Confirm that the LIF has been migrated:
network interface show -role cluster
9. Repeat step 5 on page 16 through step 8 on page 18 to migrate the LIF from port e0b to the LIF on port e0d on node 1 and
node 2.
Respond y when prompted to continue into advanced mode. The advanced mode prompt appears (*>).
12. Disable the ports to LIF clus1 and clus2 (ports e0a and e0b) on node 1 by entering the following commands:
network port modify -node node1 -port e0a,e0b -up-admin false
13. Disable the ports to LIF clus1 and clus2 (ports e0a and e0b) on node 2 by entering the following commands:
network port modify -node node2 -port e0a,e0b -up-admin false
15. Disconnect the cables that link the switches from the ports vacated in steps 3 on page 3 and 5 on page 4.
Example
In the example, the cables are disconnected from port e0a and e0b on each node and cluster traffic continues through the
switch and ports e0c and e0d on each node:
16. Cable the ports that were formerly connected to the switch to each other, creating a direct connection between the
controllers.
Example
In the following example, e0a and e0b on node1 are connected to their respective ports, e0a and e0b, on node2:
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 19
17. Change to the advanced privilege level:
set -privilege advanced
Respond y when prompted to continue into advanced mode. The advanced mode prompt appears (*>).
21. Check that communication between the nodes is good and that no RPC failures appear:
network port show
Speed (Mbps)
Node Port IPspace Broadcast Domain Link MTU Admin/Oper
------ --------- ------------ ---------------- ----- ------- ------------
node1
.
.
e0a Cluster Cluster up 9000 auto/10000
e0b Cluster Cluster up 9000 auto/10000
e0c Cluster Cluster up 9000 auto/10000
e0d Cluster Cluster up 9000 auto/10000
node2
.
.
e0a Cluster Cluster up 9000 auto/10000
e0b Cluster Cluster up 9000 auto/10000
e0c Cluster Cluster up 9000 auto/10000
e0d Cluster Cluster up 9000 auto/10000
Traffic on these ports can now travel through the back-to-back connection.
Example
The following example shows the command and illustrates the LIF revert:
network interface revert –vserver Cluster -lif node1_clus1
network interface revert –vserver Cluster -lif node2_clus1
network interface revert –vserver Cluster -lif node1_clus2
network interface revert –vserver Cluster -lif node2_clus2
24. Check the status of the nodes in the cluster to make sure health and eligibility are true:
Example
The command produces output similar to the following:
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 21
Node Health Eligibility
----- ------- ------------
node1 true true
node2 true true
25. How you proceed depends on the result of the previous command:
If... Then...
All nodes show true for both health Proceed to the next step.
and eligibility.
Any node shows false for health or Correct the problem and proceed to the next step.
eligibility.
Local Discovered
Node Port Device Interface Platform
----------- ------ ------------------------- ---------------- ----------------
node1
.
.
e0a node2 e0a FASxxxx
e0b node2 e0b FASxxxx
e0c switch_name 0/1 CN1610
e0d switch_name 0/2 CN1610
node2
.
.
e0a node1 e0a FASxxxx
e0b node1 e0b FASxxxx
e0c switch_name 0/3 CN1610
e0d switch_name 0/4 CN1610
27. Select one of the cluster ports on the first node and migrate the interfaces from that port to the other cluster port on the same
node:
network interface migrate –vserver vserver-name -lif lif-name –dest-node destination-node-
name –dest-port port-name
Example
The following illustration shows the migration of cluster LIFs from port e0c on node1 to port e0a:
network interface migrate –vserver Cluster -lif node1_clus3 –dest-node node1 –dest-port e0a
29. On the second node, select the corresponding cluster port and migrate interfaces from that port to another cluster port on the
same node:
network interface migrate –vserver vserver-name -lif lif-name –dest-node destination-node-
name –dest-port port-name
Example
The following command and illustration shows the migration of LIFs from port e0c on node2 to port e0a:
network interface migrate –vserver Cluster -lif node2_clus3 –dest-node node2 –dest-port e0a
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 23
30. Confirm that the LIF has been migrated:
network interface show -role cluster
31. Repeat step 27 on page 22 through step 29 on page 23to migrate the LIF from port e0d to the LIF on port e0b on node 1 and
node 2.
You can respond y when prompted to continue into advanced mode. The advanced mode prompt appears (*>).
34. Disable the ports to LIF clus3 and clus4 (ports e0c and e0d) on node 1 by entering the following commands:
network port modify -node node1 -port e0c,e0d -up-admin false
35. Disable the ports to LIF clus3 and clus4 (ports e0c and e0d) on node 2 by entering the following commands:
network port modify -node node2 -port e0c,e0d -up-admin false
37. Disconnect the cables that link the switches to the ports that no longer have traffic.
Example
In the example, cluster traffic continues through the direct connection between the e0a and e0b ports:
38. Cable the ports that were formerly connected to the switch to each other, creating a direct connection between the
controllers.
Example
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 25
set -privilege advanced
Respond y when prompted to continue into advanced mode. The advanced mode prompt appears (*>).
42. Revert the migrated LIFs, LIF clus3 and LIF clus4, back to their home ports:
network interface revert –vserver Cluster -lif LIF-name
Traffic on these ports can now travel through the back-to-back connection.
Example
The following example shows the command and illustrates the LIF revert:
network interface revert –vserver Cluster -lif node1_clus3
network interface revert –vserver Cluster -lif node2_clus3
network interface revert –vserver Cluster -lif node1_clus4
network interface revert –vserver Cluster -lif node2_clus4
44. Check that communication between the nodes is good and that no RPC failures appear:
network port show
Speed (Mbps)
Node Port IPspace Broadcast Domain Link MTU Admin/Oper
------ --------- ------------ ---------------- ----- ------- ------------
node1
.
.
e0a Cluster Cluster up 9000 auto/10000
e0b Cluster Cluster up 9000 auto/10000
e0c Cluster Cluster up 9000 auto/10000
e0d Cluster Cluster up 9000 auto/10000
node2
.
.
e0a Cluster Cluster up 9000 auto/10000
e0b Cluster Cluster up 9000 auto/10000
e0c Cluster Cluster up 9000 auto/10000
e0d Cluster Cluster up 9000 auto/10000
45. Check the status of the nodes in the cluster to make sure health and eligibility are true:
Example
The command produces output similar to the following:
46. How you proceed depends on the result of the previous command:
If... Then...
All nodes show true for both health Proceed to the next step.
and eligibility.
Any node shows false for health or Correct the problem and proceed to the next step.
eligibility.
Local Discovered
Node Port Device Interface Platform
----------- ------ ------------------------- ---------------- ----------------
node1
Transitioning to a two-node switchless cluster (systems with four redundant cluster-network connections) 27
.
.
e0a node2 e0a FASxxxx
e0b node2 e0b FASxxxx
e0c node2 e0c FASxxxx
e0d node2 e0d FASxxxx
node2
.
.
e0a node1 e0a FASxxxx
e0b node1 e0b FASxxxx
e0c node1 e0c FASxxxx
e0d node1 e0d FASxxxx
48. Reenable the –auto-revert parameter on each of the cluster LIFs on the cluster ports on both nodes:
network interface modify -vserver Cluster -lif LIF-name -auto-revert true
Example
In the following example, we take the steps to reenable auto-revert on the LIFs:
49. Check the cluster status of the nodes by entering the following command at the system console of either node:
cluster show
50. Verify that the cluster port LIFs are operating correctly:
network interface show
51. Verify cluster connectivity with the cluster ping-cluster local command:
You can respond y when prompted to continue into advanced mode. The advanced mode prompt appears (*>).
The command output should show connectivity between all cluster ports.
c. Return to the admin privilege level:
set -privilege admin
Copyright information
Copyright © 1994–2014 NetApp, Inc. All rights reserved. Printed in the U.S.
No part of this document covered by copyright may be reproduced in any form or by any means—graphic, electronic, or
mechanical, including photocopying, recording, taping, or storage in an electronic retrieval system—without prior written
permission of the copyright owner.
Software derived from copyrighted NetApp material is subject to the following license and disclaimer:
THIS SOFTWARE IS PROVIDED BY NETAPP "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
PARTICULAR PURPOSE, WHICH ARE HEREBY DISCLAIMED. IN NO EVENT SHALL NETAPP BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY
WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
NetApp reserves the right to change any products described herein at any time, and without notice. NetApp assumes no
responsibility or liability arising from the use of products described herein, except as expressly agreed to in writing by NetApp.
The use or purchase of this product does not convey a license under any patent rights, trademark rights, or any other intellectual
property rights of NetApp.
The product described in this manual may be protected by one or more U.S. patents, foreign patents, or pending applications.
RESTRICTED RIGHTS LEGEND: Use, duplication, or disclosure by the government is subject to restrictions as set forth in
subparagraph (c)(1)(ii) of the Rights in Technical Data and Computer Software clause at DFARS 252.277-7103 (October 1988)
and FAR 52-227-19 (June 1987).
Trademark information
NetApp, the NetApp logo, Go Further, Faster, ASUP, AutoSupport, Campaign Express, Cloud ONTAP, clustered Data
ONTAP, Customer Fitness, Data ONTAP, DataMotion, Fitness, Flash Accel, Flash Cache, Flash Pool, FlashRay, FlexArray,
FlexCache, FlexClone, FlexPod, FlexScale, FlexShare, FlexVol, FPolicy, GetSuccessful, LockVault, Manage ONTAP, Mars,
MetroCluster, MultiStore, NetApp Insight, OnCommand, ONTAP, ONTAPI, RAID DP, SANtricity, SecureShare, Simplicity,
Simulate ONTAP, Snap Creator, SnapCopy, SnapDrive, SnapIntegrator, SnapLock, SnapManager, SnapMirror, SnapMover,
SnapProtect, SnapRestore, Snapshot, SnapValidator, SnapVault, StorageGRID, Tech OnTap, Unbound Cloud, and WAFL are
trademarks or registered trademarks of NetApp, Inc., in the United States, and/or other countries. A current list of NetApp
trademarks is available on the web at http://www.netapp.com/us/legal/netapptmlist.aspx.
Cisco and the Cisco logo are trademarks of Cisco in the U.S. and other countries. All other brands or products are trademarks or
registered trademarks of their respective holders and should be treated as such.
Copyright information 29
How to send your comments
You can help us to improve the quality of our documentation by sending us your feedback.
Your feedback is important in helping us to provide the most accurate and high-quality information. If you have suggestions for
improving this document, send us your comments by email to doccomments@netapp.com. To help us direct your comments to
the correct division, include in the subject line the product name, version, and operating system.
You can also contact us in the following ways: