How to to change Storage Processor (SP) IP addresses on an integrated/unified system Changing IP addresses of CLARiiON storage processors (SPs) for integrated Celerras using proxy_ARP
Need to change CLARiiON SP's IP address
This statement does not apply: Feature: Celerra Network Address Translation (NAT) This statement does not apply: Product: Celerra Gateway Series
Product: Celerra integrated series
Feature: Celerra Proxy_ARP (Proxy ARP)
Product: Celerra NS20
Product: Celerra NS20FC
Product: Celerra NS40
Product: Celerra NS40FC
Product: Celerra NS-120
Product: Celerra NS-480
Product: Celerra NS-960
Product: Celerra NX4
Product: CLARiiON CX Series
Purpose of this solution is to describe a methodology for easily changing the SP IP addresses on Celerra Integrated platforms running NAS 5.5.31 or later, or any version of NAS 5.6 and 6.0.
Note: The storage processors (SPs) are systematically rebooted during this operation.
Current generation integrated Celerras use Proxy ARP instead of NAT by default.
With this, the /nas/sbin/clariion_mgmt tool was introduced, which can be used to change the public IP addresses for the CLARiiON storage processors.
Here are the steps to change the public CLARiiON SP addresses. In the following example, SP A's IP address is changed from xx.xx.xx.6 to xx.xx.xx.10, and SP B's from xx.xx.xx.7 to xx.xx.xx.11.
Make sure Proxy ARP is running and both SPs can be contacted using their current IP addresses.
Example:
$ /nas/sbin/clariion_mgmt -info Public IP address for SPA: xx.xx.xx.6 Public IP address for SPB: xx.xx.xx.7 Start on boot : yes Current implementation : Proxy-ARP Status : Started done
$ /nas/sbin/navicli -h xx.xx.xx.6 getagent
and
$ /nas/sbin/navicli -h xx.xx.xx.7 getagent
should return output showing Agent Revision, CLARiiON Model information, and the like.
Select new IP addresses to assign as the two new public IP addresses for the SPs. Make sure that these IP addresses are unused.
Su to root and use this command to change the SP IP addresses.
Note: in later versions of code, the capitalization of the command has changed to /nas/sbin/clariion_mgmt -modify [ -spa_ip ] [ -spb_ip ] $ /nas/sbin/clariion_mgmt -modify -spa_ip xx.xx.xx.10 -spb_ip xx.xx.xx.11 Checking if running as root...yes Checking if model is supported...yes Checking for integrated system...yes Checking if interface eth3 is configured...yes Checking if interface eth3:1 is configured...no Checking if interface eth3:2 is configured...no Checking if SP (xx.xx.xx.6) is up...yes Checking if SP (xx.xx.xx.7) is up...yes Checking if a gateway is setup for eth3...yes Step 1/14: Setting up Proxy ARP for SPA on Control Station Adding host specific route for SPA Adding rules to allow outbound traffic from SPA Adding ARP entry for SPA Updating /etc/hosts entry for SPA Step 2/14: Changing SPA IP address. Changing SPA IP from xx.xx.xx.6 to xx.xx.xx.10 (subnetmask 255.255.254.0, gateway xx.xx.xx.1) Step 3/14: Waiting for SPA to reboot. Waiting for SPA to go down................done Step 4/14: Waiting for SPA to boot up. Waiting for SPA to come back up.....................done Step 5/14: Waiting for CLARiiON software to start on SPA. Waiting for CLARiiON software to start on SPA..... Step 6/14: Updating NAS database with SPA IP address. Adding rules to allow outbound traffic from SPB Updating SYMAPI database with new CLARiiON IP addresses...done done Step 7/14: Removing old Proxy ARP setup for SPA on Control Station Removing host specific route for SPA Removing rules that allow outbound traffic from SPA Removing ARP entry for SPA Step 8/14: Setting up Proxy ARP for SPB on Control Station Adding host specific route for SPB Adding rules to allow outbound traffic from SPB Adding ARP entry for SPB Updating /etc/hosts entry for SPB Step 9/14: Changing SPB IP address. Changing SPB IP from xx.xx.xx.7 to xx.xx.xx.11 (subnetmask 255.255.254.0, gateway xx.xx.xx.1) Step 10/14: Waiting for SPB to reboot. Waiting for SPB to go down..........done Step 11/14: Waiting for SPB to boot up. Waiting for SPB to come back up.....................done Step 12/14: Waiting for CLARiiON software to start on SPB..... Step 13/14: Updating NAS database with SPB IP address. Updating SYMAPI database with new CLARiiON IP addresses...done done Step 14/14: Removing old Proxy ARP setup for SPB on Control Station Removing host specific route for SPB Removing rules that allow outbound traffic from SPB Removing ARP entry for SPB done
Ensure both SPs are up and running and can be contacted/pinged under the new public IP addresses:
$ /nas/sbin/navicli -h xx.xx.xx.10 getagent
and
$ /nas/sbin/navicli -h xx.xx.xx.11 getagent
should return output showing Agent Revision, CLARiiON Model information, and the like.
Caution! The above procedure should ideally be done during off-peak hours or a maintenance window. Although no impact is expected, this approach provides additional safety in case any problems arise during the IP address reassignment.
"How to recover proxy_arp after the customer changed SP IP addresses on CLARiiON directly"
ID:
Knowledgebase Solution
Question:
How to recover proxy_arp after the customer changed SP IP addresses on CLARiiON directly
Environment:
Product: Celerra NS Integrated series with FC option
Problem:
NS-XX FC has been setup with proxy_arp, that is, the CLARiiON storage processors (SPs) are configured with customer public IP addresses.
The customer wanted to change these SP IP addresses and performs the IP address change on the SPs directly, from Navisphere Manager or NaviCLI. After this change has been done, the SPs are not longer reachable from the control station.
This means, the CLARiiON can't be managed from the Control Station and the IP addresses can't be changed from the Control Station because they are in a different IP network and can't be reached. Furthermore, the customer changed the Control Station IP network into the same new network the SPs are in. Since the SPs are now in a different IP network and cabled internally to the Control Station NICs, there is no apparent way to get to the SPs via network to communicate with them.
Normally you would use the /nas/sbin/clariion_mgmt command to make changes to the SP IP addresses. Once the SP IP addresses have been changed on the SPs directly, the command will fail because the SPs are not reachable anymore.
# /nas/sbin/clariion_mgmt -stop Checking if running as root...yes Checking if model is supported...yes Checking for integrated system...no Checking if interface eth3 is configured...yes Checking if SP (10.64.248.224) is up...no Checking if SP (10.64.248.225) is up...no Error 4: Not an integrated system
# nas_storage -c -a
Discovering storage (may take several minutes) Error 3501: Storage API code=3593: SYMAPI_C_CLARIION_LOAD_ERROR An error occurred while data was being loaded from a CLARiiON
# /nas/sbin/clariion_mgmt -info Public IP address for SPA: 10.64.248.224 Public IP address for SPB: 10.64.248.225 Start on boot : yes Current implementation : Proxy-ARP Status : Started done
So the IP range 192.168.123.0/24 refers to the new network of the SPs and the CS eth3 interface, the 10.64.248.0/23 to your old network in that solution.
Scenario 1: The customer intended to change everything into the new IP network so you need to make modifications to the Control Station configuration to reflect these changes and re-establish communication with the SPs/proxy_arp functionality.
1. If the Control Station IP address has already been changed to the new IP network, skip to Step 3. Otherwise, go to Step 2. Use /sbin/ifconfig eth3 to check.
2. Add an IP alias to eth3 with an IP address in the new IP network. As root, type:
# /sbin/ifconfig eth3:1 192.168.123.30 netmask 255.255.255.0 eth3 Link encap:Ethernet HWaddr 00:1B:21:14:15:34 inet addr:10.64.248.228 Bcast:10.64.249.255 Mask:255.255.254.0 inet6 addr: fe80::21b:21ff:fe14:1534/64 Scope:Link UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 RX packets:4346575 errors:0 dropped:0 overruns:0 frame:0 TX packets:590720 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:375790958 (358.3 MiB) TX bytes:275747198 (262.9 MiB) Base address:0xc800 Memory:fc5c0000-fc5e0000 eth3:1 Link encap:Ethernet HWaddr 00:1B:21:14:15:34 inet addr:192.168.123.30 Bcast:192.168.123.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 Base address:0xc800 Memory:fc5c0000-fc5e0000 Note: Ensure the eth3:1 alias is not in use yet. Otherwise use eth3:2 or eth3:3
4. You should now be able to communicate with both SPs. Try to ping them and perform a navicli getagent command:
#ping 192.168.123.31 PING 192.168.123.31 (192.168.123.31) 56(84) bytes of data. 64 bytes from 192.168.123.31: icmp_seq=0 ttl=128 time=0.987 ms 64 bytes from 192.168.123.31: icmp_seq=1 ttl=128 time=0.895 ms 64 bytes from 192.168.123.31: icmp_seq=2 ttl=128 time=0.860 ms --- 192.168.123.31 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2000ms rtt min/avg/max/mdev = 0.860/0.914/0.987/0.053 ms, pipe 2
# ping 192.168.123.32 PING 192.168.123.32 (192.168.123.32) 56(84) bytes of data. 64 bytes from 192.168.123.32: icmp_seq=0 ttl=128 time=0.448 ms 64 bytes from 192.168.123.32: icmp_seq=1 ttl=128 time=0.275 ms 64 bytes from 192.168.123.32: icmp_seq=2 ttl=128 time=0.291 ms --- 192.168.123.32 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 1999ms rtt min/avg/max/mdev = 0.275/0.338/0.448/0.078 ms, pipe 2 #/nas/sbin/navicli -h 192.168.123.31 getagent Agent Rev: 6.26.0 (2.19) Name: K10 Desc: Node: A-CK200081500405 Physical Node: K10 Signature: 2142646 Peer Signature: 2142628 Revision: 3.26.10.5.010 SCSI Id: 0 Model: CX3-10f Model Type: Rackmount Prom Rev: 3.58.00 SP Memory: 1024 Serial No: CK200081500405 SP Identifier: A Cabinet: SPE3 #/nas/sbin/navicli -h 192.168.123.32 getagent Agent Rev: 6.26.0 (2.19) Name: K10 Desc: Node: B-CK200081500405 Physical Node: K10 Signature: 2142628 Peer Signature: 2142646 Revision: 3.26.10.5.010 SCSI Id: 0 Model: CX3-10f Model Type: Rackmount Prom Rev: 3.58.00 SP Memory: 1024 Serial No: CK200081500405 SP Identifier: B Cabinet: SPE3 Caution! You cannot use the SP hostnames (for example, SP A or A_CK200081500405 ) at that stage because the /etc/hosts file still contains the old IP addresses. If you cannot communicate with the SPs at that time in point, further investigation is needed before you can proceed.
5. Change storage system network configuration to reflect the new IP addresses:
# nas_storage -l id acl name serial_number 1 0 CK200081500405 CK200081500405
12. Check if proxy_arp is working and showing the new SP IPs:
$/nas/sbin/clariion_mgmt -info Public IP address for SPA: 192.168.123.31 Public IP address for SPB: 192.168.123.32 Start on boot : yes Current implementation : Proxy-ARP Status : Started done A reboot is not required for this.
It is important that the checksum in the clariion_mgmt.cksum file in both directories contains the correct checksum, otherwise clariion_mgmt -info will show the proxy_arp feature as not configured:
#/nas/sbin/clariion_mgmt -info Error 12: Not configured
If you see that, verify your checksum , update both checksum files, and rerun command. The requirement is that the proxy_arp was configured and started with the old IP addresses.
Verify the following kernel parameters:
#cat /proc/sys/net/ipv4/conf/eth0/proxy_arp # cat /proc/sys/net/ipv4/conf/eth2/proxy_arp # cat /proc/sys/net/ipv4/conf/eth3/proxy_arp #cat /proc/sys/net/ipv4/ip_forward All parameters should return "1." If that's not the case, as root type:
If the proxy_arp was never configured before, which may apply in some special cases,enter the following command to start the proxy_arp service:
/nas/sbin/clariion_mgmt -start_service
13. Check external access to the SPs is working as expected:
a. Ping SPs from customer workstation.
b. Open Navisphere using the new SP A IP address.
14. Do a final reboot of the Control Station. This will update your hosts file with the new SP IP addresses and remove old temporarily IP routes/aliases and add the new routes.
If you do not want to reboot, update hosts file with the new SP IP addresses and remove / add IP routes and aliases manually. As root type:
#/sbin/route delete 10.64.248.224 dev eth0 #/sbin/route delete 10.64.248.225 dev eth2 #/sbin/ifconfig eth3:1 down ( if configured above) #vi /etc/hosts #cat /etc/hosts ...
The IP address change was not intended or done by accident and you just want to revert the settings to the previous ones.
Alternative 1:
Attach to the SPs via the management network port and change the network settings (see the CLARiiON Procedure Generator on how to do that).T his requires the physical presence of a CE on site.
Alternative 2:
Perform some commands to re-establish communication to the SPs. Follow these steps:
1. If the control station IP address has already been changed to the new IP network, skip to Step 3. Otherwise proceed to Step 2. Use "/sbin/ifconfig eth3" to check.
2. Add an IP alias to eth3 with an IP address in the new IP network. As root, type:
# /sbin/ifconfig eth3:1 192.168.123.30 netmask 255.255.255.0 eth3 Link encap:Ethernet HWaddr 00:1B:21:14:15:34 inet addr:10.64.248.228 Bcast:10.64.249.255 Mask:255.255.254.0 inet6 addr: fe80::21b:21ff:fe14:1534/64 Scope:Link UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 RX packets:4346575 errors:0 dropped:0 overruns:0 frame:0 TX packets:590720 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:375790958 (358.3 MiB) TX bytes:275747198 (262.9 MiB) Base address:0xc800 Memory:fc5c0000-fc5e0000 eth3:1 Link encap:Ethernet HWaddr 00:1B:21:14:15:34 inet addr:192.168.123.30 Bcast:192.168.123.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 Base address:0xc800 Memory:fc5c0000-fc5e0000 Note: Ensure the eth3:1 alias is not in use yet. Otherwise use eth3:2 or eth3:3.
3. Add two IP routes to the SPs. As root, type:
# /sbin/route add 192.168.123.31 dev eth0 # /sbin/route add 192.168.123.32 dev eth2 # netstat -rn
4. You should now able to communicate with both SPs. Try to ping them and perform a navicli getagent command:
#ping 192.168.123.31 PING 192.168.123.31 (192.168.123.31) 56(84) bytes of data. 64 bytes from 192.168.123.31: icmp_seq=0 ttl=128 time=0.987 ms 64 bytes from 192.168.123.31: icmp_seq=1 ttl=128 time=0.895 ms 64 bytes from 192.168.123.31: icmp_seq=2 ttl=128 time=0.860 ms --- 192.168.123.31 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2000ms rtt min/avg/max/mdev = 0.860/0.914/0.987/0.053 ms, pipe 2
# ping 192.168.123.32 PING 192.168.123.32 (192.168.123.32) 56(84) bytes of data. 64 bytes from 192.168.123.32: icmp_seq=0 ttl=128 time=0.448 ms 64 bytes from 192.168.123.32: icmp_seq=1 ttl=128 time=0.275 ms 64 bytes from 192.168.123.32: icmp_seq=2 ttl=128 time=0.291 ms --- 192.168.123.32 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 1999ms rtt min/avg/max/mdev = 0.275/0.338/0.448/0.078 ms, pipe 2 #/nas/sbin/navicli -h 192.168.123.31 getagent Agent Rev: 6.26.0 (2.19) Name: K10 Desc: Node: A-CK200081500405 Physical Node: K10 Signature: 2142646 Peer Signature: 2142628 Revision: 3.26.10.5.010 SCSI Id: 0 Model: CX3-10f Model Type: Rackmount Prom Rev: 3.58.00 SP Memory: 1024 Serial No: CK200081500405 SP Identifier: A Cabinet: SPE3
Note: You can't use the SP hostnames ( for example SPA or A_CK200081500405 ) at that stage because the /etc/hosts file still contains the old IP addresses.
5. Once communication with the SPs has been restored, you can use the Lynx browser to change the SP IP addresses back to the old configuration:
# lynx 192.168.123.31/setup
Use tab to jump between the fields, enter will mark an option as active. See emc197661 for more information. Default login on integrated systems is nasadmin/nasadmin. If the customer changed that, you will need to ask what the current credentials are. The SP will reboot after the changes have been applied. Wait until the SP has completed and is available again before you make the changes to the second SP.
# lynx 192.168.123.32/setup
6. After the second SP rebooted successfully, check communication with the SPs and the functionality of the proxy_arp feature and do the usual health checks. # ping 10.64.248.224 # ping 10.64.248.225 # /nas/sbin/navicli -h 10.64.248.224 getagent # /nas/sbin/navicli -h 10.64.248.225 getagent # nas_storage -c -a # /nas/sbin/clariion_mgmt -info Public IP address for SPA: 10.64.248.224 Public IP address for SPB: 10.64.248.225 Start on boot : yes Current implementation : Proxy-ARP Status : Started done # nas_checkup Try to ping the SPs from the customer's network and to access them via Navisphere Manager.
7. If you added IP alias and routes in Step 2 and 3, do Steps 8 and 9 or 10. Otherwise skip these steps.
8. Remove the added routes. As root, type:
#/sbin/route delete 192.168.123.31 dev eth0 #/sbin/route delete 192.168.123.32 dev eth2
9. Remove the added IP alias. As root, type:
#/sbin/ifconfig eth3:1 down
10. Alternatively to Step 8 and 9 a reboot of the Control Station will remove these settings as well.
born_chen
1.8K 消息
1
2013年9月23日 08:00
LZ是要改SP的地址还是NAS的地址？
这几个链接供参考。
Changing IP addresses of CLARiiON storage processors (SPs) for integrated Celerras using proxy_ARP
This statement does not apply: Product: Celerra Gateway Series
Product: Celerra integrated series
Feature: Celerra Proxy_ARP (Proxy ARP)
Product: Celerra NS20
Product: Celerra NS20FC
Product: Celerra NS40
Product: Celerra NS40FC
Product: Celerra NS-120
Product: Celerra NS-480
Product: Celerra NS-960
Product: Celerra NX4
Product: CLARiiON CX Series
Note: The storage processors (SPs) are systematically rebooted during this operation.
Current generation integrated Celerras use Proxy ARP instead of NAT by default.
With this, the /nas/sbin/clariion_mgmt tool was introduced, which can be used to change the public IP addresses for the CLARiiON storage processors.
Here are the steps to change the public CLARiiON SP addresses. In the following example, SP A's IP address is changed from xx.xx.xx.6 to xx.xx.xx.10, and SP B's from xx.xx.xx.7 to xx.xx.xx.11.
Example:
$ /nas/sbin/clariion_mgmt -info
Public IP address for SPA: xx.xx.xx.6
Public IP address for SPB: xx.xx.xx.7
Start on boot : yes
Current implementation : Proxy-ARP
Status : Started
done
$ /nas/sbin/navicli -h xx.xx.xx.6 getagent
and
$ /nas/sbin/navicli -h xx.xx.xx.7 getagent
should return output showing Agent Revision, CLARiiON Model information, and the like.
/nas/sbin/clariion_mgmt -modify [ -spa_ip ] [ -spb_ip ]
Note: in later versions of code, the capitalization of the command has changed to
/nas/sbin/clariion_mgmt -modify [ -spa_ip ] [ -spb_ip ]
$ /nas/sbin/clariion_mgmt -modify -spa_ip xx.xx.xx.10 -spb_ip xx.xx.xx.11
Checking if running as root...yes
Checking if model is supported...yes
Checking for integrated system...yes
Checking if interface eth3 is configured...yes
Checking if interface eth3:1 is configured...no
Checking if interface eth3:2 is configured...no
Checking if SP (xx.xx.xx.6) is up...yes
Checking if SP (xx.xx.xx.7) is up...yes
Checking if a gateway is setup for eth3...yes
Step 1/14: Setting up Proxy ARP for SPA on Control Station
Adding host specific route for SPA
Adding rules to allow outbound traffic from SPA
Adding ARP entry for SPA
Updating /etc/hosts entry for SPA
Step 2/14: Changing SPA IP address.
Changing SPA IP from xx.xx.xx.6 to xx.xx.xx.10 (subnetmask 255.255.254.0, gateway xx.xx.xx.1)
Step 3/14: Waiting for SPA to reboot.
Waiting for SPA to go down................done
Step 4/14: Waiting for SPA to boot up.
Waiting for SPA to come back up.....................done
Step 5/14: Waiting for CLARiiON software to start on SPA.
Waiting for CLARiiON software to start on SPA.....
Step 6/14: Updating NAS database with SPA IP address.
Adding rules to allow outbound traffic from SPB
Updating SYMAPI database with new CLARiiON IP addresses...done
done
Step 7/14: Removing old Proxy ARP setup for SPA on Control Station
Removing host specific route for SPA
Removing rules that allow outbound traffic from SPA
Removing ARP entry for SPA
Step 8/14: Setting up Proxy ARP for SPB on Control Station
Adding host specific route for SPB
Adding rules to allow outbound traffic from SPB
Adding ARP entry for SPB
Updating /etc/hosts entry for SPB
Step 9/14: Changing SPB IP address.
Changing SPB IP from xx.xx.xx.7 to xx.xx.xx.11 (subnetmask 255.255.254.0, gateway xx.xx.xx.1)
Step 10/14: Waiting for SPB to reboot.
Waiting for SPB to go down..........done
Step 11/14: Waiting for SPB to boot up.
Waiting for SPB to come back up.....................done
Step 12/14: Waiting for CLARiiON software to start on SPB.....
Step 13/14: Updating NAS database with SPB IP address.
Updating SYMAPI database with new CLARiiON IP addresses...done
done
Step 14/14: Removing old Proxy ARP setup for SPB on Control Station
Removing host specific route for SPB
Removing rules that allow outbound traffic from SPB
Removing ARP entry for SPB
done
$ /nas/sbin/navicli -h xx.xx.xx.10 getagent
and
$ /nas/sbin/navicli -h xx.xx.xx.11 getagent
should return output showing Agent Revision, CLARiiON Model information, and the like.
born_chen
1.8K 消息
0
2013年9月23日 08:00
还有一个链接也可以参考：
Solution: How to modify the SP IP addresses on operational NS20FC/NS40FC and NS20/NS40 systems
https://emc--c.na5.visual.force.com/apex/KB_HowTo?id=kA0700000004I33
cxemc
362 消息
0
2013年9月23日 08:00
图形化不可以。
比如CS是管理你两个sp的，一旦图形化你把SP的IP改了，这个时候根本没法管理SP了，不通了。
只能命令行改，因为这是nas设备。改了很多次了。图形化会失败的
iamgoust
42 消息
0
2013年9月23日 08:00
这么复杂怎么看的懂呢？有没有简便方法？
比如图形化可以更改呢？
cxemc
362 消息
1
2013年9月23日 08:00
1、在线没法更改IP，必须重启，而且重启多次。若可以担当此风险就可以更改了。
2、更改起来很麻烦，按照以下步骤就好了。
EMC Knowledgebase
"How to recover proxy_arp after the customer changed SP IP addresses on CLARiiON directly"
ID:
Knowledgebase Solution
Question:
How to recover proxy_arp after the customer changed SP IP addresses on CLARiiON directly
Environment:
Product: Celerra NS Integrated series with FC option
Problem:
NS-XX FC has been setup with proxy_arp, that is, the CLARiiON storage processors (SPs) are configured with customer public IP addresses.
The customer wanted to change these SP IP addresses and performs the IP address change on the SPs directly, from Navisphere Manager or NaviCLI. After this change has been done, the SPs are not longer reachable from the control station.
This means, the CLARiiON can't be managed from the Control Station and the IP addresses can't be changed from the Control Station because they are in a different IP network and can't be reached. Furthermore, the customer changed the Control Station IP network into the same new network the SPs are in. Since the SPs are now in a different IP network and cabled internally to the Control Station NICs, there is no apparent way to get to the SPs via network to communicate with them.
Normally you would use the /nas/sbin/clariion_mgmt command to make changes to the SP IP addresses. Once the SP IP addresses have been changed on the SPs directly, the command will fail because the SPs are not reachable anymore.
# /nas/sbin/clariion_mgmt -stop
Checking if running as root...yes
Checking if model is supported...yes
Checking for integrated system...no
Checking if interface eth3 is configured...yes
Checking if SP (10.64.248.224) is up...no
Checking if SP (10.64.248.225) is up...no
Error 4: Not an integrated system
# nas_storage -c -a
Discovering storage (may take several minutes)
Error 3501: Storage API code=3593: SYMAPI_C_CLARIION_LOAD_ERROR
An error occurred while data was being loaded from a CLARiiON
Fix:
Sample configuration:
Old network:
CS eth3: 10.64.248.228
SPA : 10.64.248.224
SPB: 10.64.248.225
New network:
CS eth3: 192.168.123.30
SPA : 192.168.123.31
SPB: 192.168.123.32
IP Routing table:
# netstat -rn
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
10.64.248.225 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
10.64.248.224 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
128.221.253.201 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
128.221.253.102 0.0.0.0 255.255.255.254 U 0 0 0 eth2
128.221.253.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
128.221.252.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
192.168.123.0 0.0.0.0 255.255.255.0 U 0 0 0 eth3
128.221.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth2
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth3
0.0.0.0 192.168.123.29 0.0.0.0 UG 0 0 0 eth3
Proxy ARP:
# /nas/sbin/clariion_mgmt -info
Public IP address for SPA: 10.64.248.224
Public IP address for SPB: 10.64.248.225
Start on boot : yes
Current implementation : Proxy-ARP
Status : Started
done
So the IP range 192.168.123.0/24 refers to the new network of the SPs and the CS eth3 interface, the 10.64.248.0/23 to your old network in that solution.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Scenario 1:
The customer intended to change everything into the new IP network so you need to make modifications to the Control Station configuration to reflect these changes and re-establish communication with the SPs/proxy_arp functionality.
# /sbin/ifconfig eth3:1 192.168.123.30 netmask 255.255.255.0
eth3 Link encap:Ethernet HWaddr 00:1B:21:14:15:34
inet addr:10.64.248.228 Bcast:10.64.249.255 Mask:255.255.254.0
inet6 addr: fe80::21b:21ff:fe14:1534/64 Scope:Link
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:4346575 errors:0 dropped:0 overruns:0 frame:0
TX packets:590720 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
RX bytes:375790958 (358.3 MiB) TX bytes:275747198 (262.9 MiB)
Base address:0xc800 Memory:fc5c0000-fc5e0000
eth3:1 Link encap:Ethernet HWaddr 00:1B:21:14:15:34
inet addr:192.168.123.30 Bcast:192.168.123.255 Mask:255.255.255.0
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
Base address:0xc800 Memory:fc5c0000-fc5e0000
Note: Ensure the eth3:1 alias is not in use yet. Otherwise use eth3:2 or eth3:3
# /sbin/route add 192.168.123.31 dev eth0
# /sbin/route add 192.168.123.32 dev eth2
# netstat -rn
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
192.168.123.31 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
10.64.248.225 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
10.64.248.224 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
128.221.253.201 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
192.168.123.32 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
128.221.253.102 0.0.0.0 255.255.255.254 U 0 0 0 eth2
128.221.253.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
128.221.252.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
192.168.123.0 0.0.0.0 255.255.255.0 U 0 0 0 eth3
128.221.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth2
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth3
0.0.0.0 192.168.123.29 0.0.0.0 UG 0 0 0 eth3
# ping 192.168.123.31
PING 192.168.123.31 (192.168.123.31) 56(84) bytes of data.
64 bytes from 192.168.123.31: icmp_seq=0 ttl=128 time=0.987 ms
64 bytes from 192.168.123.31: icmp_seq=1 ttl=128 time=0.895 ms
64 bytes from 192.168.123.31: icmp_seq=2 ttl=128 time=0.860 ms
--- 192.168.123.31 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2000ms
rtt min/avg/max/mdev = 0.860/0.914/0.987/0.053 ms, pipe 2
# ping 192.168.123.32
PING 192.168.123.32 (192.168.123.32) 56(84) bytes of data.
64 bytes from 192.168.123.32: icmp_seq=0 ttl=128 time=0.448 ms
64 bytes from 192.168.123.32: icmp_seq=1 ttl=128 time=0.275 ms
64 bytes from 192.168.123.32: icmp_seq=2 ttl=128 time=0.291 ms
--- 192.168.123.32 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1999ms
rtt min/avg/max/mdev = 0.275/0.338/0.448/0.078 ms, pipe 2
# /nas/sbin/navicli -h 192.168.123.31 getagent
Agent Rev: 6.26.0 (2.19)
Name: K10
Desc:
Node: A-CK200081500405
Physical Node: K10
Signature: 2142646
Peer Signature: 2142628
Revision: 3.26.10.5.010
SCSI Id: 0
Model: CX3-10f
Model Type: Rackmount
Prom Rev: 3.58.00
SP Memory: 1024
Serial No: CK200081500405
SP Identifier: A
Cabinet: SPE3
# /nas/sbin/navicli -h 192.168.123.32 getagent
Agent Rev: 6.26.0 (2.19)
Name: K10
Desc:
Node: B-CK200081500405
Physical Node: K10
Signature: 2142628
Peer Signature: 2142646
Revision: 3.26.10.5.010
SCSI Id: 0
Model: CX3-10f
Model Type: Rackmount
Prom Rev: 3.58.00
SP Memory: 1024
Serial No: CK200081500405
SP Identifier: B
Cabinet: SPE3
Caution! You cannot use the SP hostnames (for example, SP A or A_CK200081500405 ) at that stage because the /etc/hosts file still contains the old IP addresses.
If you cannot communicate with the SPs at that time in point, further investigation is needed before you can proceed.
# nas_storage -l
id acl name serial_number
1 0 CK200081500405 CK200081500405
# nas_storage -m id=1 -network -spa 192.168.123.31 -spb 192.168.123.32
# nas_storage -c -a
Discovering storage (may take several minutes)
done
Now you are going to recover the proxy_arp functionality:
/nas/site/clariion_mgmt.cfg
/nas/site/clariion_mgmt.cksum
/nas/site/sp_info
/etc/clariion_mgmt.cfg
/etc/clariion_mgmt.cksum
$ cp /nas/site/clariion_mgmt.cfg /nas/site/clariion_mgmt.cfg.backup
$ cp /nas/site/clariion_mgmt.cksum /nas/site/clariion_mgmt.cksum.backup
$ cp /nas/site/sp_info /nas/site/sp_info.backup
Su to root, then:
# cp /etc/clariion_mgmt.cfg /etc/clariion_mgmt.cfg.backup
# cp /etc/clariion_mgmt.cksum /etc/clariion_mgmt.cksum.backup
# vi /nas/site/clariion_mgmt.cfg
# cat /nas/site/clariion_mgmt.cfg
SPA_PUBLIC_IP=192.168.123.31
SPB_PUBLIC_IP=192.168.123.32
ONBOOT=yes
CURRENT_IMPL=Proxy-ARP
# vi /nas/site/sp_info
# cat /nas/site/sp_info
192.168.123.31 A_CK200081500405 SPA # CLARiiON SP
192.168.123.32 B_CK200081500405 SPB # CLARiiON SP
# cat /nas/site/clariion_mgmt.cksum
f94074d7b4e801ad4f2827b75d71a2bc
$ md5sum /nas/site/clariion_mgmt.cfg
a78f2ba3d428504cc819906f1f1e8384 /nas/site/clariion_mgmt.cfg
# vi /nas/site/clariion_mgmt.cksum
Enter the blue marked checksum and save the file.Verify it:
# cat /nas/site/clariion_mgmt.cksum
a78f2ba3d428504cc819906f1f1e8384
# cp /nas/site/clariion_mgmt.cfg /etc/
# cp /nas/site/clariion_mgmt.cksum /etc/
$ /nas/sbin/clariion_mgmt -info
Public IP address for SPA: 192.168.123.31
Public IP address for SPB: 192.168.123.32
Start on boot : yes
Current implementation : Proxy-ARP
Status : Started
done
A reboot is not required for this.
It is important that the checksum in the clariion_mgmt.cksum file in both directories contains the correct checksum, otherwise clariion_mgmt -info will show the proxy_arp feature as not configured:
#/nas/sbin/clariion_mgmt -info
Error 12: Not configured
If you see that, verify your checksum , update both checksum files, and rerun command. The requirement is that the proxy_arp was configured and started with the old IP addresses.
Verify the following kernel parameters:
# cat /proc/sys/net/ipv4/conf/eth0/proxy_arp
# cat /proc/sys/net/ipv4/conf/eth2/proxy_arp
# cat /proc/sys/net/ipv4/conf/eth3/proxy_arp
# cat /proc/sys/net/ipv4/ip_forward
All parameters should return "1." If that's not the case, as root type:
#echo 1 > /proc/sys/net/ipv4/conf/eth3/proxy_arp
# echo 1 > /proc/sys/net/ipv4/conf/eth0/proxy_arp
# echo 1 > /proc/sys/net/ipv4/conf/eth2/proxy_arp
# echo 1 > /proc/sys/net/ipv4/ip_forward
If the proxy_arp was never configured before, which may apply in some special cases, enter the following command to start the proxy_arp service:
/nas/sbin/clariion_mgmt -start_service
a. Ping SPs from customer workstation.
b. Open Navisphere using the new SP A IP address.
If you do not want to reboot, update hosts file with the new SP IP addresses and remove / add IP routes and aliases manually. As root type:
# /sbin/route delete 10.64.248.224 dev eth0
# /sbin/route delete 10.64.248.225 dev eth2
# /sbin/ifconfig eth3:1 down ( if configured above)
# vi /etc/hosts
# cat /etc/hosts
...
192.168.123.30 ns20rdc ns20rdc
192.168.123.31 A_CK200081500405 SPA # CLARiiON SP
192.168.123.32 B_CK200081500405 SPB # CLARiiON SP
------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Scenario 2:
The IP address change was not intended or done by accident and you just want to revert the settings to the previous ones.
Alternative 1:
Attach to the SPs via the management network port and change the network settings (see the CLARiiON Procedure Generator on how to do that).T his requires the physical presence of a CE on site.
Alternative 2:
Perform some commands to re-establish communication to the SPs. Follow these steps:
# /sbin/ifconfig eth3:1 192.168.123.30 netmask 255.255.255.0
eth3 Link encap:Ethernet HWaddr 00:1B:21:14:15:34
inet addr:10.64.248.228 Bcast:10.64.249.255 Mask:255.255.254.0
inet6 addr: fe80::21b:21ff:fe14:1534/64 Scope:Link
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:4346575 errors:0 dropped:0 overruns:0 frame:0
TX packets:590720 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
RX bytes:375790958 (358.3 MiB) TX bytes:275747198 (262.9 MiB)
Base address:0xc800 Memory:fc5c0000-fc5e0000
eth3:1 Link encap:Ethernet HWaddr 00:1B:21:14:15:34
inet addr:192.168.123.30 Bcast:192.168.123.255 Mask:255.255.255.0
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
Base address:0xc800 Memory:fc5c0000-fc5e0000
Note: Ensure the eth3:1 alias is not in use yet. Otherwise use eth3:2 or eth3:3.
# /sbin/route add 192.168.123.31 dev eth0
# /sbin/route add 192.168.123.32 dev eth2
# netstat -rn
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
10.64.248.225 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
10.64.248.224 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
128.221.253.201 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
128.221.253.102 0.0.0.0 255.255.255.254 U 0 0 0 eth2
128.221.253.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
128.221.252.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
192.168.123.0 0.0.0.0 255.255.255.0 U 0 0 0 eth3
128.221.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth2
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth3
0.0.0.0 192.168.123.29 0.0.0.0 UG 0 0 0 eth3
# netstat -rn
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
192.168.123.31 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
10.64.248.225 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
10.64.248.224 0.0.0.0 255.255.255.255 UH 0 0 0 eth0
128.221.253.201 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
192.168.123.32 0.0.0.0 255.255.255.255 UH 0 0 0 eth2
128.221.253.102 0.0.0.0 255.255.255.254 U 0 0 0 eth2
128.221.253.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
128.221.252.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
192.168.123.0 0.0.0.0 255.255.255.0 U 0 0 0 eth3
128.221.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth2
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth3
0.0.0.0 192.168.123.29 0.0.0.0 UG 0 0 0 eth3
# ping 192.168.123.31
PING 192.168.123.31 (192.168.123.31) 56(84) bytes of data.
64 bytes from 192.168.123.31: icmp_seq=0 ttl=128 time=0.987 ms
64 bytes from 192.168.123.31: icmp_seq=1 ttl=128 time=0.895 ms
64 bytes from 192.168.123.31: icmp_seq=2 ttl=128 time=0.860 ms
--- 192.168.123.31 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2000ms
rtt min/avg/max/mdev = 0.860/0.914/0.987/0.053 ms, pipe 2
# ping 192.168.123.32
PING 192.168.123.32 (192.168.123.32) 56(84) bytes of data.
64 bytes from 192.168.123.32: icmp_seq=0 ttl=128 time=0.448 ms
64 bytes from 192.168.123.32: icmp_seq=1 ttl=128 time=0.275 ms
64 bytes from 192.168.123.32: icmp_seq=2 ttl=128 time=0.291 ms
--- 192.168.123.32 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1999ms
rtt min/avg/max/mdev = 0.275/0.338/0.448/0.078 ms, pipe 2
# /nas/sbin/navicli -h 192.168.123.31 getagent
Agent Rev: 6.26.0 (2.19)
Name: K10
Desc:
Node: A-CK200081500405
Physical Node: K10
Signature: 2142646
Peer Signature: 2142628
Revision: 3.26.10.5.010
SCSI Id: 0
Model: CX3-10f
Model Type: Rackmount
Prom Rev: 3.58.00
SP Memory: 1024
Serial No: CK200081500405
SP Identifier: A
Cabinet: SPE3
Note: You can't use the SP hostnames ( for example SPA or A_CK200081500405 ) at that stage because the /etc/hosts file still contains the old IP addresses.
# lynx 192.168.123.31/setup
Use tab to jump between the fields, enter will mark an option as active. See emc197661 for more information. Default login on integrated systems is nasadmin/nasadmin. If the customer changed that, you will need to ask what the current credentials are. The SP will reboot after the changes have been applied. Wait until the SP has completed and is available again before you make the changes to the second SP.
# lynx 192.168.123.32/setup
# ping 10.64.248.224
# ping 10.64.248.225
# /nas/sbin/navicli -h 10.64.248.224 getagent
# /nas/sbin/navicli -h 10.64.248.225 getagent
# nas_storage -c -a
# /nas/sbin/clariion_mgmt -info
Public IP address for SPA: 10.64.248.224
Public IP address for SPB: 10.64.248.225
Start on boot : yes
Current implementation : Proxy-ARP
Status : Started
done
# nas_checkup
Try to ping the SPs from the customer's network and to access them via Navisphere Manager.
#/sbin/route delete 192.168.123.31 dev eth0
# /sbin/route delete 192.168.123.32 dev eth2
#/sbin/ifconfig eth3:1 down
Jeffey1
2.8K 消息
1
2013年9月23日 22:00
楼主，根据你提供的消息我给几点建议：
1、NS480修改IP地址不能在线操作，服务肯定会有中端，建议联系专业技术支持工程师进行修改。
2、从楼主提供的系统版本号（29）来看，这个是后端block的版本号，前端nas版本号一般都是5.6、6.0或者7.0之类。修改IP地址一般前端nas和后端block都需要修改，步骤非常多，详细参考楼上兄弟提供的操作文档。