Life After Production: Equallogic PS100s

We did some major SAN upgrades over the summer and due to this, found ourselves with (2) extra PS100 chassis that we took out of production. With Dell EOLing these units we did not want to put them into any kind of critical/production capacity BUT we also didn’t want to get rid of 2 perfectly good units just because of support status either, these units still have A LOT of life left in them!

Our original units were configured for 14x250GB drives leaving very little usable space for our purposes. We knew we would need to re-drive these units for them to be usable but to what, Dell quit making the upgrade kits years ago and the highest available drives at the time (750GB I believe) still wouldn’t cut it. After having some “on the side” conversations with some Dell Engineers and folks we heard rumors of these units working fine with 1+TB hard drives.  DISCLAIMER: Doing ANYTHING “non-Dell” will void your warranty, OH WAIT, it doesn’t matter anymore anyway 🙂

Did some searching and found some nice 2TB Western Digital Enterprise-Class Drives on Newegg, we decided on these units: http://www.newegg.com/Product/Product.aspx?Item=N82E16822236522
5 year warranty is nice since Western Digital is now our RMA source, 24/7 rated, etc.
I’m sure other drives would work fine but these are enterprise-rated which is my personal preference, but your mileage may vary. We ordered 14 of these 2TB units, pulled the remaining 250GB drives, took about 30 minutes to swap out all the drives, 4 quick screws on each drive sled and bam, DONE! Now the moment of truth, would the array recognize these units and work properly?

Short answer YES, you can see the screen below, all drives were recognized at their correct size.

EQL1

Depending on how you RAID this you will get some of the rough usable options below:

EQlR50

EQLR10

EQLR6

We chose RAID10 which gives us around 10.6TB usable space which is plenty for our initial use, the nice thing with RAID10 is you can move to 50 or 6 if more capacity is needed down the road. For our purposes this unit will be primarily doing backup/archive work and NVR recordings from our security cameras.

We still have an additional unit that we can use for a parts depot or for other projects down the road. Even though these units are older/EOL, there are still lots of parts from refurb groups, Ebay, etc. out there so we are not too worried, plus with the 5 year drive warranty we will be covered for quite a while down the road, PLUS these units when they were in production were rock-solid, in 5+ years of use we never had one single MAJOR hardware/software issue minus the occasional drive failure!

Originally our units were running 5.2.2 code which was EXTREMELY reliable, I confirmed the drives were recognized/worked fine at this version before updating to 6.0.7 which also is working fine.

On a side note, you really have to watch Dell’s support wording, based on my understanding you can not join these older units to your production storage group, all members of a group must be current/in-support for things to be considered VALID. In our situation this was not a big deal, these units will be completely separate from our Production Equallogics and have no production workloads on them BUT make sure if you do some DIY work on your older units to keep them separate, don’t want to risk support on your production units!

Advertisements

Author: Travis Kensil

Director of IT. Husband and father. Michigan beachbum.

7 thoughts on “Life After Production: Equallogic PS100s”

  1. A question about these arrays: I have a pair that Dell will not take my money to support, but they have a nasty interoperability bug in firmware <=6.0.6 that can corrupt vmfs volumes.

    I don't have any production-grade EQL gear that would entitle me to downloading fw from Dell; any chance you can spare a 6.0.8 firmware?

  2. I was told by a Dell Support person while working on another issue that anything over 1TB in those Sans will start corrupting when you have filled them to a percentage of the capacity.

    I don’t know what that percentage is, but apparently according to him there isn’t enough address space on the controller to address that much storage and it “Wraps” and starts overwriting the already used sectors of the hard drives.

    I’ve googled it and not found anyone else reporting this phenomenon. Can you confirm that you have filled your arrays and not had a problem reading the data back?

  3. Old thread but I’m re-driving the same array. We have it up running with 1T drives in an R10. I’m slowly replacing each with 1.5 T drives. By slow I mean replace a drive then let it rebuild… which is agonizingly slow. Since I don’t have any volumes created would you suggest just powering it off, plugging new drives in and see what happens?

  4. kw,
    That sounds possible but we have not yet encountered any issues. We are using about 8TB currently and have around 2TB free and have not yet encountered any issues. We also don’t have anything of production value that if we did have an issue we would have problems either.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s