How do tow node failures to opretative ScaleIO Configuration?

What version of ScaleIO, and what OSes, revisions, etc? you Say Windows SDC, what were the MDM/SDSes?

How many networks for data traffic, and what speeds?

Dear Rush.

Thanks reply.

Infomation.

ScaleIO version 2.0

and SDC OS is Windows 2012 R2 (SDC version 2.0)

MDM (2 node) is Windows 2012 R2 Core and ScaleIO version 2.0

SDS Node is Multiple OS, 5 node CentOS 6.X (last update) and 4 node Windows 2012R2 Core

SDS ScaleIO version 2.0

Base Network is 1G switch (24 Port switch)

and All node (with SDC) is 2 x 1G NIC.

When iometer test start (in SDC), i see 20 ~ 23 MB/s  (800 ~ 830 IOPs)

Configuration appaently there be problems?

(about Fault-set and node config)

Best Regards

Yoon

Yoon,

Unless you have very few (or very slow) disks per server, the limitation here should be the 1gbit network.

We don't recommend 1gbit network, even with 2 NICs.  Multiple 10gbit NICs are recommended.

Since you've mentioned fault sets, how many do you have?

What are the disks like? Any SSD, or all rotating? How many?

Have you configured IP roles?

Dear Rush.

Sorry, I'm update late.

I looked back to peruse the UserGuide.

And, as I've found the problem.

I config Fault Set

1st Fault Set (SDS1, SDS2,SDS3 include)

2nd Fualt Set (SDS4, SDS5, SDS6 include)

3rd Fualt Set (SDS7, SDS8, SDS9 inlcude)

Before, failure was occurred to SDS1 and SDS4

>> I see problem with SDC volume.

But, This time the failure occurred on the SDS1, SDS2

>> no problem with SDC volume

I think, I had misunderstood the meaning of the Fault Set.

I understand, SDS in two or more faulures in the same Fault Set, SDC volume has no problem.

Do you think, I test collect?

or I understand is still wrong?

Best Regards
yoon

HI,sdyoon

I Have a question.when use fault set in scaleio.is it occupied storage capacity or take additional overhead for system management?thanks.