PS6100x and PS4000 Firmware

Things about to remember when working with EQL

1. All EQL members within the group must run on the same FW.  Only during update you run a mixed setup for a short time. So update existing PS4000 first and than add PS6100 which than have to be upgraded

2. To get support all members need active maintenance within the group. So check the support for your ps4000

3. Your are not allowed to update to a newer FW on a member without active maintenance

If you speak about vSphere svMotion it doesnt matter if your placed your member to a existing pool, a different one or a in a 2nd. group. (We leave VAAI and xcopy out of the game here). But there is a easier way.....

If all members in the same pool the volumes a split automaticly over all members if you like. If you vacuate a old member out of the pool the data will migrate automaticly. Its easy and fast.

If the members are in different pools you can move one EQL volume after another between the pool. Its easy but need some clicks and you can only migrate one volume at once.

So.. ... there is no svMotion needed and no downtime of course.

Regards,

Joerg

Hi Don -

So upgrade the 4000' firmware from 6.0.5 to 7.1.4?

Then upgrade the 6100 to 7.1.4?

Then proceed with the rest of your instructions?

Correct?

Can you explain - FX2 doesn't have enough capacity for both iSCSI and LAN? We built it with 2 I/O Modules with 8 Ports each, 4x Broadcom 1GB NICs and 2x CMCs

Sorry bit of a newbie.

Thanks!!

Hi,

well...most of the time you have to provide diags from the GroupLeader and the effected member.  Maybe you can see the problem when GL run on an unsupportet member within your group. Also its most likely the case that the GL runs on the oldest, because it was the first member in your group.

About the FW thing.. its the same as with vSphere SnS and lot of others. If your SnS short ends after the Release of vSphere 5.0 youre not allowed to download and run 5.1 or 5.5 on your hosts. It doenst matter if the LIC Keys havent change Your entitlement stops with 5.0..

IIRC there is a statement about the PS5000 doesnt support the up comming 8.x and what that means for you. Same is true about unsupportet members.

Maybe Don or one of the others can say something about.

We collect our members which runs out of support in their own E-O-L group

.

About not selling support for our PS6010X after 4.5 years we ask Dell "again" after the their first "no". They sell us 2 more years.

Regards,

Joerg

Joerg - thanks for the additional info.  Actually, my GL is not the oldest member in the group and in any case, so what if it happens to be the GL which actually generates the diags for the group, as long as the issue I'm calling in about doesn't originate from the unsupported array.

Don - sorry, but, Dell should have been upfront about this policy instead of waiting until an array actually goes EOL.  How come the sales people don't make this stuff clear when they come in to do their dog and pony show?  I was told the arrays we were purchasing would ALWAYS remain compatible with future EQL models.

Besides the CYA policy from Dell, there is no technical reason to have to put the EOL array into its own group.  It's already in its own storage pool within the group and will never mix with supported arrays.  It's an unnecessary waste of staff time and resources to do this kind of reconfiguration.

Hi Don - Can you expand on the process of - RAIDset is built and all the 6100X interfaces are enabled, then move it into the same pool with the PS4000.  Then you can move that PS4000 to that temp pool?

Do you have a document or link for this process and how to accomplish this?

Thank You!

He talks about the standard install and setup procedere. But instead doing this when the member is already part of your production Pool you do all these steps in a "maintanence pool". I our case this is always the "default" one.

So add it to the "maintanence" pool and take time to finshed

- Raid "re"build

- Upgrade PS FW and Drive FW

- Bringing up all ETHs

- Create documentation

-  Create a Test volume and see if the Hosts can connect to

- Run some benchmarks

- Perform tests like a controller failover (The "restart" option within the Group Manager GUI)

and then.... move it into your production pool. The  volume loadbalacing for distributing the data is enabled by default on the group. So as soon as the member joins the pool it becomes active (we dont talk about Raid preference here).

Regards,

Joerg

ETHs stands for?

:) Thanks,

ETHernet.

Your EQL have eth0, eth1 .... and so on.  With the integrated setup wizzard you only setup one ethernet device which is most likly eth0.

I have seen a lot of installation where the Admin have forget to configure the other devices :)

Regards,

Joerg

Hi Don - When you move the 6100x to the same Pool as the PS4000, is that when the data gets "moved, copied or replicated" to the PS4000? If there is no need to use vMotion, how does the data get transferred? to the new array? Or does this automatically happen when you move the new array into the same Pool and the PS4000?  

Check the advanced tab for the group within the Group Manager. There is an option called "Enable load balancing in pool". Its enable by default.

So if you have more than one member in a pool a Volume spans over all of them.  In theorie it goes over up to three members but it can be more than three.

This loadbalancing is how EQLs scales out.

Regards,

Joerg

Ah right.

I remember that we have to disable page moving/balacing in the tech support mode during trouble shooting.

@RMB07

Click on a volume in the GroupManager and you will see the distribution.

The volume spreads over 3 EQLs.

Regards,
Joerg