Why defined storage matters? Sergey Goncharov Solution Architect, [email protected]

AGENDA

● Storage and Datacenter evolution

● Red Hat Storage portfolio

● Red Hat Storage

● Red Hat Storage

● Red Hat Storage roadmap and future

● Q/A

WHAT IS SOFTWARE-DEFINED STORAGE?

SERVER-BASED CENTRALIZED CONTROL OPEN ECOSYSTEM

3 THE DATACENTER IS EVOLVING

DEVELOPMENT APPLICATION DEPLOYMENT APPLICATION STORAGE MODEL ARCHITECTURE AND PACKAGING INFRASTRUCTURE

Waterfall Monolithic Bare Metal Data Center Scale Up

Agile N-Tier Virtual Services Hosted Scale Out

DEVOPS MICROSERVICES CONTAINERS HYBRID CLOUD SOFTWARE-DEFINED STORAGE

4 SERVER-BASED STORAGE

Server-based storage is the use of software and standard hardware to provide services traditionally provided by single-purpose storage appliances (similar to server virtualization, which uses software to emulate servers), providing increased agility and efficiency.

USER USER USER USER USER USER

Appliance Appliance Appliance Distributed Cluster of Services

5 STORAGE ORCHESTRATION

Storage orchestration is the ability to provision, grow, shrink, and decommission storage resources on-demand and programmatically, providing increased control and integration of storage into a software-defined data center.

Web Console API Command Line A browser interface designed A full API for automation and A robust, scriptable command-line for integration with outside interface for expert operators managing distributed storage systems

PROVISION INSTALL CONFIGURE TUNE MONITOR

Full life cycle management for distributed, software-defined data services

6 GROWING INNOVATION COMMUNITIES

97 AUTHORS/MO

• Contributions from Intel, SanDisk, CERN, and Yahoo 2,453 COMMITS/MO

• Presenting Ceph Days in cities around the world and quarterly virtual Ceph Developer Summit events 260 POSTERS/MO

33 AUTHORS/MO

• Over 11M downloads in the last 12 months 97 COMMITS/MO

• Increased development velocity, authorship, and discussion has resulted in rapid feature expansion 138 POSTERS/MO

7 THE FUTURE OF STORAGE

TRADITIONAL STORAGE OPEN, SOFTWARE-DEFINED STORAGE Complex proprietary silos Standardized, unified, open platforms

USER USER USER USER ADMIN

O

p

S

e Control Plane (API, GUI)

o

n

f

t

S

w

ADMIN ADMIN o ADMIN a

u

r

e

r

e Ceph Gluster +++

Custom GUI Custom GUI Custom GUI e d r

Proprietary Hardware Proprietary Hardware Proprietary Hardware r a a Standard Computers d w n d r

a and Disks t a S

Proprietary Proprietary Proprietary H Software Software Software

8 THE RED HAT STORAGE PORTFOLIO

Share-nothing, scale-out architecture provides E C E durability and adapts to changing demands R R U A O W Ceph Gluster S T

F management management N O E S

P Self-managing and self-healing features O Ceph data service Gluster data service reduce operational overhead

Standards-based interfaces and full APIs ease integration with applications and E D R

R systems A A W D D N R A T A S H Supported by the experts at Red Hat

9 HOW STORAGE FITS

RED HAT STORAGE

PHYSICAL VIRTUALIZATION PRIVATE CLOUD CONTAINERS PUBLIC CLOUD

10 RED HAT GLUSTER STORAGE , software-defined storage for unstructured file data at petabyte scale

Media, video Machine, Log GeoSpatial Persistent Documents Data Storage RED HAT GLUSTER STORAGE ADVANTAGES

OPEN • Based on GlusterFS open source community project Open, software-defined • Uses proven local (XFS) distributed file and object store • Data is stored in native format

• Uses an elastic hashing algorithm for data placement SCALABLE No Metadata Server • Uses local filesystem’s xattrs to store metadata • Nothing shared scale-out architecture

• Global name space ACCESSIBLE Multi-Protocol the Same • NFS, SMB, object, HDFS, Gluster native protocol Data • Posix compliant

• GlusterFS is based on (FUSE) MODULAR • Modular stackable arch allows easy addition of features No Kernel Dependencies ...without being tied to any kernel versioniant

ALWAYS-ON • Synchronous with self-healing for server High-Availability across data, failure systems and applications • Asynchronous geo-replication for site failure GLUSTER ARCHITECTURE Distributed scale out storage using industry standard hardware

NFS CIFS FUSE

SERVERS WITH LOCAL DISKS

Aggregates systems to one cohesive unit and presents using common protocols WHY RED HAT CEPH STORAGE FOR OBJECTS?

• Proven at Web-Scale for • Flexible Storage for Your Applications • Data Protection, Reliability & Availability Enterprises Demand • Open, Community Based Software Defined Storage for Object Storage • Cost Effective Object Storage to Help You Maximize Your Storage Budget CEPH ARCHITECTURE

APPLICATION HOST / VM CLIENT

RGW RBD CEPHFS*

A web services gateway for A reliable, fully-distributed A distributed file system object storage, compatible block device with cloud with POSIX semantics & with S3 & Swift platform integration scale-out metadata

LIBRADOS

A library allowing apps to directly access RADOS (C, C++, Java, Python, Ruby, PHP)

RADOS

A software-based, reliable, autonomous, distributed object store comprised of self-healing, self-managing, intelligent storage nodes and lightweight monitors

* TECH PREVIEW

15 NEW INTEGRATED MANAGEMENT WITH RED HAT STORAGE CONSOLE 2

ADMINS CLIENTS • Easy to use graphical interface to manage storage cluster life cycle RED HAT STORAGE CONSOLE 2

-based deployment tools for RADOS. RGW, RBD, CALAMARI installation, importation, and granular configuration from CLI or GUI • Monitoring and troubleshooting with statistical and graphical information about cluster components CLUSTER OF STANDARD SERVERS AND MEDIA ROAD MAP: RED HAT GLUSTER STORAGE

TODAY (v3.1) V3.2 (H1-2016) FUTURE (v4.0 and beyond)

ERASURE CODING COMPRESSION E E INODE QUOTAS E BIT ROT DETECTION DEDUPLICATION R R FASTER SELF-HEAL R TIERING HIGHLY SCALABLE CTRL PLANE O O CONTROLLED REBALANCE O

C SNAP SCHEDULE C C NEXT-GEN REPLICATION DISTRIBUTION E E SMB 3 E ACTIVE/ACTIVE NFSV4 L L (ADVANCED FEATURES) L pNFS I SMB 3 (BASIC SUBSET) I I MULTI-CHANNEL F F F

SELINUX C C C QOS AT-REST ENCRYPTION E SSL ENCRYPTION E E CLIENT SIDE CACHING

S ( IN-FLIGHT) S S

DEVICE MANAGEMENT T T T M GEO-REPLICATION, M DYNAMIC PROVISIONING M NEW UI

G SNAPSHOTS G OF RESOURCES G GLUSTER REST API

M DASHBOARD M M

Gluster 3.7, RHEL 6, 7 Gluster 3.8, RHEL 6, 7 Gluster 4, RHEL 7 RED HAT CEPH STORAGE 2 (today)

OBJECT BLOCK CORE MANAGEMENT • Support for global clusters • Mirroring • Bluestore (tech preview) • Red Hat Storage Console 2 • AD, LDAP, Keystone v3 • Performance optimizations • Lifecycle management • Expanded API compatibility • Ansible support • NFS gateway (tech preview)

RED HAT CEPH STORAGE 2.1 (Q4 2016)

OBJECT BLOCK • AWS Secure Token Service • iSCSI (STS)

UPSTREAM ACTIVITY (in the future)

OBJECT BLOCK CORE FILE • Metadata search • Bluestore general availability • Active/active metadata • Tiering to public cloud • Distributed client-side cache • Data checksums servers • Object-level encryption • Improved iSCSI support • End-to-end SSL • Snapshots • Compression • Improved multi-threading • integration