You're
already
using
CEPH-based
storage cluster,
and would like to
analyze and optimize your systems
professionally?
Or consider the use of
Software Defined
Storage solutions
with CEPH
to offer
your
own highly
scalable storage and
cloud solutions,
and seek expert advice
/
consulting?
Then you are at
the right place: it concepts Oliver Liebel - your Expert/Specialist
for
HA-Storage-Cluster-Systems and Software Defined Storage.
Whats this Thing about Ceph? - A short Story around a
large Scalable System...
Traditional Storage Cluster already encounter their limits in terms of
scalability and flexibility.
The rapidly increasing data growth in almost all areas of IT
is far beyond scope of the scalability of
traditional hardware-based storage.
And - No end
so far:
All operating worldwide, renowned market research institutes assume
that global data volume will reach more than 40,000 exabytes(!)
up
to
the Year
2020.
Regarding
studies of the IDC, the majority of companies
is not prepared to manage this flood of data.
All of these Companies have to deal with rapidly growing data volumes. If they
can't, they'll have to handle significant financial losses.
Unlike proprietary, hardware-based storage solutions,
CEPH provides an open, software-defined storage platform ( "Software
Defined Storage"), which allows almost unlimited scalability of
Storage with physical, virtual and
cloud-based resources.
It provides tremendous cost savings, avoiding the typical vendor
lock-in, and also allows unlimited capacity
expansions.
The open source, scale-out storage solution CEPH blurs the
boundaries of "stale" Storage far behind, and will
give you 3 ways do deliver your Storage to your Business or to your
Customers:
- CephFS: the highly scalable, Self Healing
Cluster-Filesystem for direct Use
- RBD: The Rados Block Device - multi purpose
high Scalable and Self-Healing Blockdevice, that can act, e.g as a
Backend Device for general Storage Purposes, or as a direct KVM/libvirt
VM-Backend, without the need of a Filesystem (see below)
- Rados Object Store / Rados Gateway:
Build your own, massively Scalable Cloud Storage, directly on top of
your Ceph-Scale Out Storage Cluster Pools.
CEPH clusters can be expanded with new nodes and/or
Disks at runtime, no special hardware is
required,
Geo-Site Cluster between data center are daily
business,
and to make the best and most effective Use of your
Storage Hardware, Ceph has Hot-/Cold-Storage policy-based Features already
builtin and implemented, to reduce the Costs per TB, and give you the Redundancy
you need.
CRUSH-Maps (Controlled Replication under
Scalable Hashing) gives you the greatest flexibility, to adjust the
Ceph-Storage Layout according to your real Hardware Layout: e.g.
From Clients view, Data-Chunks are splitted on datacenter Level,
then down to room-level, rack, nodes, Disks - you've got full
control down to the smallest chunk of Data.
Ceph-clusters are Self-Healing - Data on
broken OSDs (the Ceph Disks/"Workhorses") will automatically
rebuild on other OSDs.
Deployment
of virtual machines raises a
new Level:
Through
RADOS Block Device (RBD), virtual machines
can be deployed (no need for a cluster FS for Live Migration)
directly on a RBD-Image.
With
RADOS Gateway,
CEPH Cluster can provide
S3 / Swift
compatible cloud storage, and traditional claims to a cluster FS are
handled
with CephFS.
Moreover,
NFS / iSCSI
on top of
CEPH / RBD can
be
perfecty combinded with the
Reliability and Intelligence of Industry proven Pacemaker Cluster
Systems,
to ensure maximum HA to your Storage Based Applications.
CEPH, by far the most scalable Software Defined Storage solution
of the World, is already in Use
by many Enterprise Scale
production environments,
with high reliability and great success.
Almost all Linux Enterprise Cloud Solutions (SUSE / Red Hat) are now
based on CEPH.
[click
here to get back]
|