28 Posts

October 16th, 2014 06:00

Hi Bill,

Jyothi is correct, in that fail-safe networking is configured on the VNXe storage systems by default. However, this is only true for the VNXe3100, VNXe3150, and VNXe3300 models. This is noted in the following VNXe High Availability white paper for these models.

VNXe High Availability - VNXe3100, VNXe3150, and VNXe3300

The VNXe3200 does not use fail-safe networking. In the example you provided, the NAS Server that is associated with Ethernet Port 2 on SPA would remain on SPA until the SP reboots or is placed into Service Mode. To survive a link failure on SPA Ethernet Port 2, your NAS Server would need to leverage Link Aggregation. Please take a look at the VNXe3200 High Availability white paper for more information.

VNXe3200 High Availability

October 15th, 2014 02:00

Hi Bill,

Fail Safe Networking(FSN) is configured on the VNXe storage systems by default.If one of the ports in the storage system fails,FSN will automatically reroute the I/O internally to the corresponding physical port on the peer storage processor.For more information you can refer to the document mentioned below(page 30)

EMC VNXe SERIES STORAGE SYSTEMS

Thanks

Jyothi

2 Intern

 • 

143 Posts

October 15th, 2014 06:00

Thanks Jyothi, that is helpful.  I knew this was built in to the VNXe 3100, 3150, and 3300, but I was not sure if it was also built in to the VNXe 3200.  Glad to hear it is. 

Does anyone know the answer to my questions regarding MAC addresses, failover time, etc.?

2 Intern

 • 

143 Posts

October 16th, 2014 08:00

Charles,

Thanks so much for taking the time to provide a correct answer on this for the VNXe 3200.  I figured that the VNXe 3200 was completely different since it is no longer using the old DART 5.6 code base.  I looked through the VNXe 3200 High Availability white paper, and it clearly explains how to set up LACP for network high availability.  The biggest issue I see, however, is for customers without stackable switches.  Most non-stackable switches do not support LACP across switches, so it would not be possible to set up LACP across SP A Eth4 and Eth5 to two different (non-stackable) switches.  In contrast, the old "Fail Safe Networking" capability in the VNXe 3100, 3150, 3300, and Celerra/VNX Unified models, works fine across non-stackable switches, since it is essentially active/passive and not using LACP.  Given this constraint, if a potential VNXe 3200 customer wishes to have true network high availability with the VNXe 3200 (on CIFS/NFS), is it a fair statement to make that stackable switches are mandatory (or at the very least, switches that support "virtual stacking")?

Thanks,

Bill

28 Posts

October 17th, 2014 12:00

Yes, the non-stackable switch configuration does not provide the same level of network high availability on the VNXe3200 as it once did with the other VNXe models (VNXe3100, VNXe3150, and VNXe3300).

2 Intern

 • 

143 Posts

October 20th, 2014 08:00

Thanks!  I really appreciate the great help in comparing the VNXe 3200 to the older generation of VNXe.  Makes perfect sense.

No Events found!

Top