Honestly, best bet is to just use the current version of Virtual Storage Console anyway....It puts the best current settings for the combination (Filer OS, 7-mode/cDOT, ESXi version) of stuff that you are running.

--tmac

*Tim McCarthy* *Principal Consultant*

On Mon, Dec 29, 2014 at 3:47 PM, Philbert Rupkins <philbertrupkins@gmail.com

wrote:

I believe I found part of the reason for the conflict. Despite addressing vSphere 5 later in the technical report, TR3749 states the following in the Document Roadmap section on page 6:

"This document, TR-3749, is the main document in a set of documents from NetApp covering virtualization on NetApp, specifically with VMware products. This document is specifically in reference to VMware vSphere 4."

With that being said, I'm pretty comfortable moving forward with the vSphere 5.5 recommendations in the KB article rather than the Netapp Storage Best Practices for VMware vSphere.

On Mon, Dec 29, 2014 at 11:50 AM, Philbert Rupkins < philbertrupkins@gmail.com> wrote:

...

Toasters,

It appears the following documents conflict with one another when it comes to recommendations for the value of Net.TcpipHeapMax in ESXi.

---

*How to configure VMware vSphere 5.x for Data ONTAP 7.3 and 8.x* https://kb.netapp.com/support/index?page=content&id=1013275

Net.TcpipHeapMax

For vSphere 5.0/5.1 set 128

For vSphere 5.5 set 512 *TR3749 - NetApp Storage Best Practices for VMWare vSphere* http://www.netapp.com/us/media/tr-3749.pdf

"While the NetApp best practice remains to set Net.TcpipHeapMax at the maximum value of 128MB, tests have shown ESXi 5 to be less sensitive to depletion of TCP/IP heap memory."

---

It seems that TR3749 may be out of date (last update 2011) as it doesnt distinguish between vSphere 5.0/5.1 and 5.5. To my knowledge, the 128MB maximum TCP/IP heap size is applicable to vsphere 5.0 and 5.1. I believe 5.5 increases this maximum to 512MB.

I assume the current recommendation for vSphere 5.5 against a filer running ONTAP 8.1.4 7-Mode is 512MB. Can somebody confirm? Is there a best practices doc like TR3749 for vSphere 5.5 in a 7-mode environment?

Thanks, Phil

---

Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters

If you are using cDOT....it even shows if you have direct or indirect NFS/iSCSI/FCP paths If needed/possible, one could migrate a LIF (of course, you need to follow a best practice of only allowing one datastore per IP on cDOT to migrate LIFs like that!)

--tmac

*Tim McCarthy* *Principal Consultant*

On Tue, Dec 30, 2014 at 12:56 AM, Fletcher Cocquyt fcocquyt@stanford.edu wrote:

Seconded - just checked my VSC plugin in vCenter is reporting 403 errors (probably needs to be updated - why isn't that built into the plugin system?)

now oncall, I get to read https://library.netapp.com/ecmdocs/ECMP1392339/html/GUID-7E05AC55-F737-4796-...

When its working, the plugin shows best practice VMware:Netapp datastore settings compliance and via vCenter lets you update any sub-optimal settings (some will require an EXSi host reboot to take effect)

----- Original Message ----- From: "tmac" tmacmd@gmail.com To: "Philbert Rupkins" philbertrupkins@gmail.com Cc: toasters@teaparty.net Sent: Monday, December 29, 2014 1:29:45 PM Subject: Re: Conflicting Recommendations?

Honestly, best bet is to just use the current version of Virtual Storage Console anyway....It puts the best current settings for the combination (Filer OS, 7-mode/cDOT, ESXi version) of stuff that you are running.

--tmac

Tim McCarthy Principal Consultant

On Mon, Dec 29, 2014 at 3:47 PM, Philbert Rupkins < philbertrupkins@gmail.com > wrote:

I believe I found part of the reason for the conflict. Despite addressing vSphere 5 later in the technical report, TR3749 states the following in the Document Roadmap section on page 6:

"This document, TR-3749, is the main document in a set of documents from NetApp covering virtualization on NetApp, specifically with VMware products. This document is specifically in reference to VMware vSphere 4."

With that being said, I'm pretty comfortable moving forward with the vSphere 5.5 recommendations in the KB article rather than the Netapp Storage Best Practices for VMware vSphere.

On Mon, Dec 29, 2014 at 11:50 AM, Philbert Rupkins < philbertrupkins@gmail.com > wrote:

Toasters,

It appears the following documents conflict with one another when it comes to recommendations for the value of Net.TcpipHeapMax in ESXi.

---

How to configure VMware vSphere 5.x for Data ONTAP 7.3 and 8.x https://kb.netapp.com/support/index?page=content&id=1013275

Net.TcpipHeapMax

For vSphere 5.0/5.1 set 128

For vSphere 5.5 set 512 TR3749 - NetApp Storage Best Practices for VMWare vSphere http://www.netapp.com/us/media/tr-3749.pdf

"While the NetApp best practice remains to set Net.TcpipHeapMax at the maximum value of 128MB, tests have shown ESXi 5 to be less sensitive to depletion of TCP/IP heap memory."

---

It seems that TR3749 may be out of date (last update 2011) as it doesnt distinguish between vSphere 5.0/5.1 and 5.5. To my knowledge, the 128MB maximum TCP/IP heap size is applicable to vsphere 5.0 and 5.1. I believe 5.5 increases this maximum to 512MB.

I assume the current recommendation for vSphere 5.5 against a filer running ONTAP 8.1.4 7-Mode is 512MB. Can somebody confirm? Is there a best practices doc like TR3749 for vSphere 5.5 in a 7-mode environment?

Thanks, Phil

---

Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters

---

Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters