In questo episodio, sono com Jeff Lundberg (Senior Product Marketing Manager, HDS) e parliamo di:

– HDS HCP ecosystem (introduzione e componenti)
– Come vengono installati i singoli componenti
– HCP S10

Trascrizione completa dell’episodio

Enrico:                        Hi, everyone, and
welcome back to Juku Bits. Today, I’m with Jeff Lundberg, senior project
manager at HDS for object storage. Hi, Jeff. How are you?

Jeff Lundberg:           I’m doing well, Enrico. Thank you.

Enrico:                        Good. I’m quite a fan of
object storage. Your solution from the end user’s perspective is very
interesting, because you are building a kind of ecosystem around the core
platform. Can you introduce that concept, and talk about the company that you
are developing?

Jeff Lundberg:           Sure. We have an offering, just the
core object store itself is called Hitachi Content Platform. We’ve had that
for, oh, boy, eight, nine years now. What we’ve seen over that time is object
stores started in the archive and compliance space, but they’ve really become
the core storage infrastructure behind a lot of the cloud services that people
find in the market today and are gravitating towards. Some of the key use cases
for that are things like file services and file sync and share. We’ve expanded
the object store to support those use cases with some of our own offerings in
this space.

                                    In addition to the Hitachi
Content Platform as the object store, we have our own cloud gateway in the
Hitachi Data Ingestor, and we have our own file sync and share offering. That
is the Hitachi Content Platform Anywhere. Really, this gives us the ability to
bring an integrated solution to our customers, as opposed to them having to put
together different pieces of open source and third-party products and try to
maintain and support those over time.

Enrico:                        That’s really
interesting. Can you spend a few more words about the NAS gateway and the sync
and share? How do you deploy them to the customer? Are they digital appliances
or physical appliances? How-

Jeff Lundberg:           Yeah. That’s one of the nice things
about our offering. All of these components are available as appliances, where
we provide the servers and the storage and the software and the switching, and
so forth, as part of an appliance package, or we can provide just the servers
with the software so they can attach to existing storage investments that
people have made around Hitachi arrays and so forth. We also offer them as
software-only solutions that can run in a VMware environment. You can really
pick the right mix of, maybe I do want a big, integrated, tested, configured
appliance in my data centers, but when I’m dealing with my remote and branch
offices, I just want to push a VMware image out there.

                                    When I want to send a very small
physical box, like a little toaster oven to these remote and branch offices I
have or these cloud services customers, so that they have a local cache they
can access the data that they need. They have that physical box that they can
go touch and say, “All my data is in there.” When it comes to HCP
Anywhere, it’s a lot like the sync and share products that people are familiar
with – the dropboxes, the boxes, etc, but it’s managed and controlled and
deployed within your own data center. You have that extra layer of security and
control over all of this data that your users are synchronizing and sharing
across all their different devices. It supports Windows phones, iOS, Android,
Macs and PCs and so forth.

                                    It’s really trying to bring some
of those consumer-type products that have really gotten their way into the
enterprise just through users deploying them on their own, but giving
organizations an option to bring all of that in-house, have a little bit more
control over it, and have a single solution that they can then offer to all of
their lines of business or their cloud customers, and say, “You can do
this in a more controlled, centralized manner where we can apply all of our
compliance rules and do discovery and things like that.”

Enrico:                        One more thing that is
really interesting about your platform, it’s not that you are not developing
only the front end ecosystem but also the back end ecosystem. In the past, you
had the ability to go on public clouds, so tiering functionalities. Lately, you
have introduced a new appliance, if I can call that an appliance, the HCP S10.
Can you explain how it works, and which kind of role it has in the [inaudible
00:05:13] ecosystem?

Jeff Lundberg:           Absolutely. One of the things that
we’ve tried to do with the content platform is make it very flexible from the
kinds of applications and data sources that can work with it, so we support a
lot of different protocols from S3 to HTTP to Rest-type protocols, as well as
the more standard NFS and CIFS and even some of the SMTPs, if you wanted to say
journal exchange emails in. On the back end, we’ve also, traditionally, used
this with our array technologies, but we have versions that come with on-board
disk. We can tier to clouds, as you mentioned, so it’s not just, put everything
in your HCP. You can store it all there, and then you can use service plans to
automate the movement of data into Amazon or Google or Azure or any other S3
type of cloud service provider.

                                    That still gives you all that
centralized tracking of this data: where it flows in from, where it flows out
to, encrypt it before it goes to the cloud. A lot of different capabilities
that really enable that hybrid architecture. For those that want to have that
cloud price point, but don’t want to bear the risk of putting their data in the
public clouds, we did come up with the S10 that you mentioned. We
traditionally, again, used arrays as part of the storage for HCP. Now we have
an option where people can use an erasure coded storage platform that allows
them to use more of the commodity disk type price points, but get the kinds of
data protection capabilities that are needed in an enterprise setting, but
doing that at a lower price point.

                                    This, again, is all about choice.
I can have my hot data on an array or in an on-board disk. I can take some of
my colder data and put it into a local erasure coded tier. Maybe, at some point
when I do understand that there’s a data set that is perfectly fine to put out
into the public cloud, I can pick that up, store it out in the public cloud,
and still keep track of it all from my HCP.

Enrico:                        Good. Sounds like a real
interesting story. We have only five minutes, but I would like to continue for
hours talking about your server. In any case, thank you very much for your
time, Jeff.

Jeff Lundberg:           Thank you, Enrico. Happy to talk
again any time.

Enrico:                        Bye. Bye-bye.

Jeff Lundberg:           Bye.