0 Replies Latest reply: Jun 14, 2016 1:53 AM by iEMC APJ RSS

VNX Data Mover Failover and Failback

iEMC APJ

VNX Data Mover Failover and Failback

Share: image001.png

Please click here for all contents shared by us.

Introduction

 

    Data Mover is a hardware component of VNX for File. It reads the data from back-end storage by UltraFlex I/O technology to provide the data access. It supports NFS, CIFS and pNFS protocols. The component has scalability and redundancy.

    Data Mover failover/failure recovery mechanism is a capability of the system. If one of the devices failed and could not work normally, another device can automatically take over the work of the failed device.


Detailed Information

 

 

Data Mover Overview

 

 

1.png
  • X-Blade enclosure is Data Mover
  • A VNX has up to 8 Data Movers, each X-Blade enclosure has up to 2 Data Movers
  • Data Mover is something like a high-performance and high availability server
  • It is used to connect to the data and network
  • In the operating environment of VNX NAS side, a failover procedure will not affect the performance
  • Data Mover supports hot-plug
  • Support Multi-Protocol Label Switching (MPLS)
  • The failure of Control Station fails does not affect Data Mover
  • Data Mover does not have internal disks

 

 

Data Mover Type

 

  • Use the command nas_server -list to show the basic information of Data Mover information:
2.png

    Type 1 stands for the primary Data Mover, type 4 stands for the standby Data Mover

 

  • Use nas_server -a to show the detailed information of Data Mover:
3.png

 

 

Data Mover Failover

 

4.png
  • When primary Data Mover fails, the standby Data Mover will take over the work
  • Some Celerra models have only one Data Mover, in that case, it is impossible to implement failover function
  • Control Station plays a monitoring role in this procedure



    When primary Data Mover fails, there comes three situations: 

Failover Policy

                       Description

Auto

Immediately activate the standby Data Mover

Retry

Control Station first tries to recover the primary Data Mover. If the recovery fails, the Control Station activates the standby

Manual (Default)

Control Station shutdown the primary Data Mover and takes no other action.The standby must be activated manually



    Use server_standby command to failoaver the failed Data Mover to the standby one.

    Below is how to manually activate the standby Data Mover to primary:

5.png

 



RCM Data Mover Failover/Failback Procedure

 


Failover

 

a.     Use nas_server -l to show the basic information of Data Mover:

6.png

b.    Restart the standby Data Mover (server_y is the name of the Data Mover)

server_cpu server_y -reboot -monitor now

7.png

c.    Use the following command to verify the standby Data Mover. Contacted means it works normally:

/nas/sbin/getreason

8.png

d.    Switch the standby Data Mover to the primary Data Mover (server_x is the name of the device)

server_standby server_x -a mover

9.png

e.     After switching the Data Mover:

nas_server -l

10.png

 



Failback

 

a.     After the failver, make sure both Data Movers are in contacted State

/nas/sbin/getreason

11.png

b.    Once the failure is fixed, restore the primary Data Mover

server_standby server_x -r mover

12.png

c.     Make sure both Data Movers are in contacted state

/nas/sbin/getreason

13.png

d.    Both two Data Movers are back to the initial state

nas_server -l

14.png

 

    EMC recommends that each Data Mover has a standby one. Once the primary Data Mover failed, the standby Data Mover can take over the work, ensure the host applications are working normally. Therefore, the configuration of the standby Data Mover should be same as the primary Data Mover's.