Virtual walk through of adding new disks to DMX4

it's a little more involved then that. You CE will put in a CCA (change control ticket) to get this disk addition approved. During that time he will put a lock on the frame to keep the bin file consistent. During this "freeze" you will not be able to make any changes to the array (symconfigure). When the new bin file gets approved, CE will loaded it in Symmwin. Typically new drives will be created in a new disk group. At that time you can run symdev list and should see the new hypers on the bottom of your list. At that point you can start your provisioning.

Symwin is the application EMC uses to control array configuration. It's installed on a 1U server that resides in the same cabinet where all the FA/DA/Cache cards are. On older boxes it used to reside on a laptop. You and i (customers) have no access to Symwin, Solution Enabler has come a long way that we don't require as many bin file changes as we used to before. ..so much can be done from the cli (that makes changes in the bin file), but things like disk addition, FA/DA/Cache expansion all require CE to make changes in Symwin.

Thanks but...

1. What the heck is SymWin :)? I keep reading about it here...
2. Is the infamous Bin file something the customers can modify and load?

Sorry, I'm a DIYer and hate to feel dumb when the answer is 'Call your EMC rep' or 'Call your Honda dealer when the idiot light comes on'

New Storage virtual Walkthrough.
Once the disks are ordered let your CE know how many and off what type.
Lets say all the same size and speed.
All disk purchases come with one free bin file.
Let the CE know how many and what size splits you need.
The bin can be locked before delivery if you know when they are coming and as long as you are happy that you cannot make any changes to the system during this period.
The CE will add the new disks to the bin file and assign them to a disk group.
You can add these to existing disks groups and/or new ones.
This is helpful if you have tiered storage (Doesn't everyone? ;-)).
Or if you have different app's managed by disk groups.
Then he will then 'carve' the disks to your requirements, EG 128 18414 mirrored or raid splits, from the disk groups as per your instructions.
Create 4 meta volumes using 32 splits etc etc etc.
Once the bin file is complete it is put through the CCA process.
Change Control Automation is the online tool we use to ensure that all the correct guidelines have been followed before EMC makes any change to your system. It is not for the DMX family alone but all of our changes done be CE's on all our products will go through this system.
There is an automated part to the system that checks that the CE has submitted all the correct forms to accompany the change or upgrade.
In this case there would be a before and after bin file, Activity Assurance Form and Symmetrix Procedure. The system will also be aware of the changes/additions that will occur and may approve the activity automatically if the rules engines rules are all met IE code levels etc.
Most activities will also get checked by someone from the CCA group, depending on the complexity of the bin it will get checked by aa hardware RTS or a software RTS or both.
Once all the necessary approvals are completed the system e-mails the CE and he is good to go.
If you have existing DAE's drives can be inserted before the activity and then the activity can be done by the Remote Change Management team. They will dial in and load the bin and let the CE and or customer know when it is complete.
If there are DAE's to be added with the storage then the CE will come to site load the bin and add the DAE's and drives when the Symmwin scripts instructs him (Or her...) again at this point the CE could leave site and let the RCM team complete.
Once this is done all the new drives are VTOC'd you are then free to allocate/provision your new storage.
This is as basic as I can get it and be rest assured there is a lot of stuff going on behind the scenes when a CE carries out a CCA activity. The CE needs time to build the bin and CCA requires all activities to be submitted 48hrs before execution.
Our normal lead time on a request for a change is 5 days. Anything less than this does not allow us time to prepare for the activity properly to ensure that we maintain data integrity across your systems.
Hope this is what you where after SG201.

I guess ChampionG's answer covers it all. If you need additional answers, please turn off the idiot light and call your EMC rep

You need any information:
1. Physically install the disks into the DAE:
The Symmetrix DMX-4 950 can be configured in several versions
depending on the number of director-pairs required and number of
drives.
2. run symcli xyz command to make the array see the new drives:
You need the bin file for thin new upgrade.
3. create hypers using abc command:
You need the bin file.
4. create metas:
You can to use this new bin file also.

Ohh ..you are talking about that inferior NetApp company ..aren't you ? .

I appreciate the detailed reply.

However, to EMC, wow! this is archaic. I felt like I stepped back into the 70s when computers were giant iron machines with turning reels and required people with pocket protectors and taped eye glasses to look after them. We buy what we feel comfortable managing and this is not it. Afterall we're all well paid, educated and worthy of managing a storage array given the right information and tools. Wanna know how some other storage arrays add disks:

1. Rack the new disk shelf and turn it on
2. connect the fiber channel cables into the FC loop
3. verify the new disks are see using vol status -s
4. add it to an existing aggr with aggr add or create a new aggr with aggr create
5. go home feeling good that you did all this in less than 20 mins not weeks

You can be rest assured StorageGuy there is method in out madness.
Whenever a bin file is applied to a box with customer data on it the bin is checked numerous times to ensure accuracy. Also we ensure any changes we do with the new bin do not adversely affect the system or the customer data already on it.
I am sure if you had a DU/DL scenario on your DMX you would be asking why we do not have processes/procedures etc to ensure that we maintain data integrity during our upgrades. If we do have any issues during a planned upgrade all the tech people involved have a great trail of info to help with any issues and make fault finding, diagnosis and resolution much smoother and quicker than it may have been.
I have been doing this for 8+ years now and I like the system.
Keeps my customers data safe and confirms I am doing my job properly.
Archaic maybe, but it works.
I go home knowing my customer data is intact and safe.
Geoff

I was close to hitting "Reply" button a few times .. But I'm biased (you all know) thus didn't reply before .. Reading your post I'm satisfied I didn't reply before .. 'coz you explained everything I wanted to explain, simply in a better way ..

IMHO playing with disks and cables can be funny .. but only when I'm at my home. When I'm "playing" with terabytes of customer valuable data, I prefer to leave the cables at their place and ask specialists to change boards/drives/cables/whatever you want.

ThX AranH

Thanks for the compliment Stefano. I really do think it is a learning curve issue and understanding the reasoning behind the change control process. I had similar thoughts after getting our DMX4 as my experience with storage arrays was around midrange arrays where for years I could do all my own disk/shelf upgrades. Having to depend on EMC to make what before were simple changes took some adjustment on my part. But now that I am more familiar with the DMX I understand the reasoning and it is no longer an issue.

I don't understand the logic especially when I can cause more destruction by using a single symcli than connecting up disk shelves. If I expand an existing lun without setting the parameter to preserve the data there goes the data...bye bye...no warning, nothing, done. Same with the Clariion, I could right click and say delete raidgroup...gone.

I don't understand the logic of 'DMX is stable because we do the actual addition/cabling'. I thought DMX was stable on its own account not by preventing customer goofups. In other words, if I built a cheap Linux NAS and told my customers to not go near it does it make it stable. If that's the perception here then I disagree.

Furthermore, I really hate to say this but I have yet to come across an EMC CE that was well versed, they're all green behind the ears. I've had to point out to them which module/disk/component was where. Unfortunately, I'll answer my own question, it's all business and EMC support is gnerates a lot of revenue for the company...more than sales as is typically the case with most software/hardware. However, they're really shutting out the sysadmins from doing anything...hopefully others don't follow suit or we'll be waiting weeks for Dell to come out and add a second drive to our home PCs...while we sit home wondering why we bothered with a 4 year degree in computer science and 10+ years as systems administrators.

I would disagree with that logic as you can take down an entire bus if you are not careful in contrast to destroying a single logical device with the example you listed.

Requiring EMC to perform physical upgrades to expensive high end arrays does not really result in 'shutting out the sysadmins from doing anything'. Again, the vast majority of array management tasks are performed by you, the sysadmin.

If you really think a bunch of cobbled together Linux boxes are comparable to a DMX in regards to performance and stability, then I think you should spend some more time reading about the DMX platform. Honestly that is not a fair or informed statement.

I don't think EMC is that different from other vendors in regards to the field techs. There are some very good ones, there are some green ones. That is why they have detailed procedures for their techs and engineers that are directly available to those field techs.

Aran I strongly disagree with StorageGuys .. but please don't take it too personal .. . You don't work for EMC !!

However thank you for your great support and for your opinion.