Tiering up: Our man struggles to make sense of the storage landscape

All is confusion. The old certainties are gone. New certainties just don’t exist. The shifting shapes, players, products and technologies in the storage landscape are seen through fog. How the heck does everything fit together? After four days in Silicon Valley meeting startups the bewilderment ratio is even higher. It’s like …

[Elephantpm]

Thank you for a helpful article

Nice.

[Anonymous Coward]

Some guy just the other day had a rant about people calling storage systems SANs. This acronym stands for storage area networks, i.e. the stuff that connects storage systems together, and to host systems.

By getting this wrong, you end up with nonsensical sentences like this one, from the article:

"Then there are two main flavours of on-premises storage, with the first being the no-SAN SANs, the virtual SANS made by aggregating server’s direct-attached storage into a logical SAN."

And to exactly which server are you referring? Or did you mean to write " servers' " (spaces added by me to pedantically highlight the location of the apostrophe.)

[Anonymous Coward]

I think SAN was always a forced acronym. (meaning the actual storaged device(s) rather than the wires connecting them) People just like having an A for Area in the middle so it sound like it fits in with traditional LAN / WAN topology.

See also NAS - um , Network Area Storage? whassat mean?

some will say "aatached" storage - but that was just a hasty correction after they realised Area made no sense. Musch the same as RAID has changed from from inexpensive to identical

[Lusty]

RAID hasn't changed at all, although occasionally I see "Independent" rather than inexpensive. I've never seen that gibberish version of NAS either. SAN was SAN because at the start it was a storage network of FICON, FC or something similar and nothing to do with the network.

[Anonymous Coward]

NAS has always been 'attached' and RAID 'inexpensive'. I have never heard 'area' used for NAS but it would just be wrong. Some RAID does require identical devices, so this error is more understandable but still and error. The other 'I' that gets used is 'independent' as some many of the old school vendors could never be described as 'inexpensive'. The 'A' in SAN was because it might be different to the LAN/WAN, but always 'area'.. Anyone remember ESCON? iSCSI makes merging SAN and LAN possible, but be careful of contention.

[MityDK]

The only useful part of your post was where you accurately self portray as an insufferable pedant.

[Anonymous Coward]

You felt the need to reply though, didn't you, although one could hardly describe your own contribution as "useful". You did incorrectly suggest that I described myself as insufferable, which I didn't. I'm incredibly sufferable, as it turns out.

How about using the El Reg favourite, the down vote? That will make me cry.

Anyway, what bit didn't you like? The way people use network terminology to describe end points, my highlighting of the nonsensical sentence, or my pointing out of the grammar error? And for the record, I don't really generally care if people can't spell or punctuate: I know plenty of intelligent people who suffer from dyslexia. I'm just not a fan of professional journalists doing it, although of course, errors do happen and none of us is infallible.

[Discovery]

Good summary of the players Chris. Each one with a history of pumping protocols, speeds, feeds and data location recommendations focused on the 5 9's and the economics of storage. DataGravity wants to up the ante on that conversation. The infrastructure must get intelligent at all layers. Data-Aware is the new game in storage. Before that data goes anywhere we want to examine it for key attributes like sensitivity, security, ownership, dormancy etc. Essentially data gets a voice and a seat at the business table.

The next major inflection point in this industry happens when you can see what's in your data so you can take action on it before someone else does. We've witnessed some pretty amazing things when people put their data on our systems. In every single case, 100% of the time, they find sensitive information in that data that would cause business harm if it got into the wrong hands. We're finding it at the point of storage. We're showing them the Who, What, When, Where and Why of their data and they are blown away by it.

If you’re talking about the five 9’s, those are the right charts. With the five W’s, you are missing some key players.

[elan]

my 2cnts

- two kind of storage types will remain : local a/o remote

- decision differentiator is data security a/o latency

- san/nas/b2d will merge ( bad for netapp )

- bu/restore/archive will be a data service

- incumbents: ibm will leave the table, hp has nothing beyond 3par, netapp sinking ship

- remaining: emc (best management), hds (best engineers)

- new players: tegile / kaminario / oracle(not new but investing...)

- market will be divided by 4- 6 players , as it was 3y ago

- remote storage players (google/azur/aws/...) will collapse (=the next crash ) - one or two remain

[Yaron Haviv]

ignoring the Dark Horse ?

Chris,

i believe the post ignores the Dark Horse in storage, the fastest growth for cloud vendors is not in Object Storage, rather in managed data services/lakes, part of the overall trend of moving from IaaS to PaaS & SaaS, i.e. (micro) Service Centric Architectures.

AWS storage services like Dynamo, RedShift, Aurora, Kinesis, .. are growing exponentially AFAIK, just open GCP front page (cloud.google.com), out of 14 listed services 8 are Data Services (i.e. storage), one is object, and NONE match the categories listed in the post (SAN, NAS, Hyper-Converged, ..), go visit Azure and you will see the same. its only reasonable to expect this trend will get to on-premise data centers (the ones which will survive the cloud assault).

the reason we consume lots of SAN is mainly DBs (a Data Service) and VM images, DBs move to service based consumption like the ones above (using NoSQL & DAS or Data Lakes), even Oracle figured it out and is transforming, this will be re-enforced by the growth in BI/Analytics & IoT demanding much larger and faster data services, VM images will move to stateless containers to enable workload elasticity and DevOps (see my post: http://sdsblog.com/2015/09/16/cloud-native-will-shake-up-enterprise-storage/)

yes, there will be lots of legacy, people still buy FICON and mainframes, but in few years the storage landscape will be very different, the current turmoil we see is only the tip of the iceberg.

Yaron

[bobpott]

ZFS using 3DXPoint as L2ARC

Oracle or Cisco (Better equipped to deliver low latency NVMe) would come to mind to play this marginalizing not only most of AFA talks of today, but change the narrative of data life cycle and backup/archiving; harder is to guess how is secondary storage (if still explicit) landscape look like...