![Page 2: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/2.jpg)
2
2 Andreas Koll
AGENDA
● Motivation für Open & Software Defined Storage● Red Hat Storage Overview
● Red Hat Gluster Storage● Red Hat Ceph Storage
● Use Cases
![Page 3: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/3.jpg)
3
3 Andreas Koll
AGENDA
● Motivation für Open & Software Defined Storage
![Page 4: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/4.jpg)
Andreas Koll4 Andreas Koll4 Andreas Koll4
To offer a unified, open software-defined storage portfolio that delivers a range of data services for next generation workloads
thereby accelerating the transition to modern IT infrastructures.
THE RED HAT STORAGE MISSION
![Page 5: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/5.jpg)
5 Andreas Koll
● Hybrid-flash arrays
● Public cloud
● Software-defined storage
● Open source
Source: Gartner Top 10 Tech Trends 2014
New wave of storage disruption and innovation
![Page 6: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/6.jpg)
6 Andreas Koll
Source: Gartner Top 10 Tech Trends 2014
Top tech trends for 2014 IT needs to have a plan in place by 2016
1. Software-defined networking
2. Software-defined storage
3. Hybrid cloud services
4. Integrated systems
5. Application acceleration
6. The internet of things
7. Open compute project
8. Intelligent datacenter
9. IT Demand
10.Organizational entrenchment and disruption
![Page 7: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/7.jpg)
7 Andreas Koll
Cornerstone of the software-defined datacenter
Datacenter evolution
Source: IDC Report – Taxonomy for Software-Defined/Based Storage
COMPUTECOMPUTE
SOFTWARE- DEFINED / BASED
COMPUTE(Virtualization)
SOFTWARE- DEFINED / BASED
COMPUTE(Virtualization)
STORAGESTORAGE
SOFTWARE-DEFINED / BASED
STORAGE
SOFTWARE-DEFINED / BASED
STORAGE
NETWORKINGNETWORKING
SOFTWARE-DEFINED / BASED
NETWORKING
SOFTWARE-DEFINED / BASED
NETWORKING
ENVIRONMENTALENVIRONMENTAL
SOFTWARE-DEFINED / BASED
FACILITIES
SOFTWARE-DEFINED / BASED
FACILITIES
DATA CENTER FABRIC DATA CENTER FABRIC
![Page 8: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/8.jpg)
8 Andreas Koll
Traditional storage system costs
Standard x86 systems costs
Source: IDC research report: “The economics of software-based storage”
20% OPEX52% TCO5 years
37% TCO3 years
In addition, you save up to:
Up to 1/3 the cost of traditional Storage
GAP
![Page 9: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/9.jpg)
9
9 Andreas Koll
AGENDA
● Motivation für Open & Software Defined Storage● Red Hat Storage Overview
● Red Hat Gluster Storage● Red Hat Ceph Storage
● Use Cases● TCO & Customer example● Q&A
![Page 10: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/10.jpg)
10 Andreas Koll
2 Produkte.....
Red Hat Gluster Storage Red Hat Ceph Storage
10 Andreas Koll
![Page 11: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/11.jpg)
11 Andreas Koll
Red Hat Gluster Storage
● synchronous replication with self-healing for server failure● asynchronous geo-replication for site failure
● NFS, SMB, object, HDFS, Gluster native protocol● Posix compliant
● GlusterFS is based on file-system in user space (FUSE)● modular stackable architecture allows easy addition of
features without being tied to any kernel version
● uses an elastic hashing algorithm for data placement ● uses local file system’s extended attributes to store metadata● shared nothing scale-out architecture
● based on GlusterFS open source community project● uses proven local file system ( XFS : fast, performand, scalable )● data is stored in native format
MODULARNo kernel dependencies
ACCESSIBLEMulti-protocol to the same data
SCALABLENo metadata server
OPENOpen, software-defined distributed
file and object storage system
![Page 12: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/12.jpg)
12 Andreas Koll
Red Hat Storage Technology Highlights
● huge namespace● software only, Open Source Solution● runs in User Space● simple and extremely easy to set up● unified file- and object storage● modular stackable architecture (scale-out)● basefilesystem XFS in native format● no MetaData server due DHT (elastic hash algorithm)
![Page 13: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/13.jpg)
13 Andreas Koll
ENTERPRISEMOBILITY
INCREASE DATA, APPLICATION AND INFRASTRUCTURE AGILITY
CLOUD APPLICATIONS
CONVERGED COMPUTE AND STORAGE
FILE SERVICES OPEN OBJECT APIs
OPEN, SOFTWARE-DEFINED STORAGE PLATFORM
SCALE-OUT STORAGE ARCHITECTURE
PHYSICAL
Standard x86 systemsScale-out NAS solutions
VIRTUAL
Include idle or legacy resources
CLOUD
EBSEBS
BIG DATA WORKLOADS
ENTERPRISE APPLICATIONS
DATA SERVICES
13
![Page 14: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/14.jpg)
14 Andreas Koll
RED HAT STORAGEFOR ON-PREMISE
SERVER (CPU/MEM)
1TB
• global namespace
• aggregates CPU, memory,network capacity.
• deployed on Red Hat-supported servers and underlying storage: DAS, JBOD
• scale out performance and capacity as needed.
• replicate synchronouslyand asynchronously.
RED HAT STORAGE DEPLOYMENT ON-PREMISE
RED HAT STORAGEFOR ON-PREMISE
1TB
Scale out performance, capacity, and availability
Sca
le u
p c
apac
ity
huge NAMESPACE
...
...SERVER(CPU/MEM)
............ ... ...
![Page 15: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/15.jpg)
15 Andreas Koll
RED HAT STORAGE SERVER FOR PUBLIC CLOUD
RED HAT STORAGEFOR PUBLIC CLOUD
EBS
Scale out performance, capacity, and availability
Sca
le u
p c
apac
ity
• GlusterFS Amazon Machine Images (AMIs)
• The only way to achieve high availability of Elastic Block Storage (EBS)
• Multiple EBS devices pooled
• POSIX compatible (no application to rewrite required to run on Amazon EC2)
• scale out capacity and performance as needed
huge namespace
RED HAT STORAGE DEPLOYMENT ON AMAZON CLOUD
...
...EC2
............ ... ...
![Page 16: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/16.jpg)
16 Andreas Koll
CIFS HADOOP ENABLEMENT
REPLICATION
MULTI-SITE DR
MULTI-TENANT:NAMESPACE
VOLUMESNAPSHOT
CLIENT/ PRESENTATION
BACKEND/ PERSIST
SAMBA USER APP QEMU
SWIFT
MANAGE
FUSE NFS
TRANSLATORS
TRANSLATORS
GLUSTERFS FRAMEWORK
GLUSTERFS
NETWORK STACK
NETWORK DEVICE PLATFORM BLOCK DEVICE
HARDWARE ENABLEMENT
LOCAL FILESYSTEM
LOGICAL VOLUME MANAGEMENT
XFS OTHER
RED HAT ENTERPRISE LINUX
PL
AT
FO
RM
MA
NA
GE
AB
ILIT
Y
RED HAT STORAGE TECHNOLOGY STACK
![Page 17: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/17.jpg)
17 Andreas Koll
Red Hat Storage volume setup anatomy
RHSOperatingSystem
RHSOperatingSystem
Brick #2
Brick Brick
![Page 18: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/18.jpg)
18 Andreas Koll
Red Hat Storage volume connect options
RHSOperatingSystem
RHSOperatingSystem
![Page 19: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/19.jpg)
19 Andreas Koll
Distributed Volume
basic volume-type
NO high availability
distributing files across available bricks
![Page 20: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/20.jpg)
20 Andreas Koll
Replicated Volume
basic volume-type
high availability (up to 3 copies)
distributing files across available bricks
![Page 21: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/21.jpg)
21 Andreas Koll
Distributed-Replicated Volume
combined volume-type
high availability (up to 3 copies)
High performance
distributing files across available bricks
![Page 22: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/22.jpg)
22 Andreas Koll
Striped Volume (tech. preview)
datafiles striped across bricks
use in high concurrency environments w/ large files
# of bricks should equal stripe count
![Page 23: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/23.jpg)
Andreas Koll23 Andreas Koll23 Andreas Koll23
![Page 24: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/24.jpg)
24 Andreas Koll
RHS Server
Mirror Mirror Mirror
…
GlusterFS native client – data flowClients talk directly to the data bricks based on elastic hash
App server 1Running GlusterFS
client
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
![Page 25: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/25.jpg)
25 Andreas Koll
NFSAccessibility from UNIX and Linux systems
● Standard NFS v3 clients connect to GlusterFS NFS server process (user space) on storage node
● Mount GlusterFS volume from any storage node
● GlusterFS NFS server includes network lock manager (NLM) to synchronize locks across clients
● Better performance for reading many small files from a single client
● Load balancing must be managed externally
● Standard automounter is supported
![Page 26: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/26.jpg)
26 Andreas Koll
SMB/CIFSAccessibility from Windows systems
● Storage node uses Samba with winbind to connect with Active Directory environments
● Samba uses Libgfapi library to communicate directly with GlusterFS server process without going through FUSE
● SMB clients can connect to any storage node running Samba
● SMB version 2.0 supported
● Load balancing must be managed externally
● CTDB is required for Samba clustering
![Page 27: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/27.jpg)
27 Andreas Koll
Mirror Mirror Mirror
…
SMDB
SWIFTNFS
NFS, CIFS & OBJECT – data flow
IP load balancing & high availability ctdb rrdns, ucarp, haproxy, hw (f5, ace etc.)
Glusterfsd brick server
RHS Server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
Glusterfsd brick server
…
Application serverRunning
NFS/SMDB/HTTP client
5
23
1
4 Distributed – replicated volume
SMDB
SWIFTNFS
SMDB
SWIFTNFS
SMDB
SWIFTNFS
SMDB
SWIFTNFS
SMDB
SWIFTNFS
. . . .
Clients first talk to the mounted storage node and are then directed to the data bricks. Non-
native protocol adds an additional network hop.
![Page 28: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/28.jpg)
28 Andreas Koll
Object access of GlusterFS volume
● Built upon OpenStack’s Swift object storage
● GlusterFS is the back-end file system for Swift
● Implements objects as files and directories under the container
● Accounts are implemented as GlusterFS volumes
● Store and retrieve files using the REST interface
● Support integration with SWAuth and Keystone authentication service
![Page 29: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/29.jpg)
29 Andreas Koll
Object store architecture
End Users
Node 1
Node 2
Node n
memcached
Red Hat Storage Volume (swift acount)
Coantainer server
Proxy server
Object server
Account server
Load Balancerglusterfsd
![Page 30: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/30.jpg)
30 Andreas Koll
Remote site / disaster recovery
Geo-replication : global data protection and availability
SITE A
SITE B
REMOTE SITE / DR
SITE C
![Page 31: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/31.jpg)
31 Andreas Koll
Forklift upgrades Migration costs
• Moves data to new node
• Upgrade without disruption
• Easily de-commission systems
Built-in re-balancing frees up IT budget
![Page 32: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/32.jpg)
32 Andreas Koll
HA for NFS and CIFS
Any IP failover tool can work for NFS Appliance based load balancers with heartbeat such as F5
Linux heartbeat, ucarp, CTDB
Not all failover works for CIFS as that requires some session handling
CTDB (Cluster Trivial Database) is what we use It is very simple to configure
Works for NFS
Works for CIFS
Is very robust and configurable
Round robin DNS for load balancing You can use any load balancer you want
RRDNS is simple to configure and works well
Prevents hot spots of activity
![Page 33: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/33.jpg)
33 Andreas Koll
● Open rest API based on OpenStack® Swift
● Https encrypted dataover the wire
● Encrypt data at restwith DMCRYPT
● Identity and authentication support keystone, Kerberos, andRed Hat IDM
ENTERPRISEMOBILITY
CLOUD APPLICATIONS
ENTERPRISE APPLICATIONS
Global shared namespace
Rest object APIs based on Swift
Simultaneous, secure object access to file dataUnified file and object storage
![Page 34: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/34.jpg)
34 Andreas Koll
Storage co-resident applicationsConverging compute and storage
Web apps Hadoop Custom apps ● Run native Red Hat Enterprise Linux applications
● Process data locally
● Eliminate an entire tier of hardware
● Use c-groups to manage resources
● Use Red Hat Enterprise Linux KVM
![Page 35: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/35.jpg)
35 Andreas Koll
Red Hat Storage Server Hadoop connectorApply Hadoop analytics directly on production data
● Direct data access
● Eliminate HDFS ingestion overhead
● Included with your Red Hat Storage Server subscription
● Alongside or instead of HDFS SCALE-OUT OPEN SOFTWARE-DEFINED STORAGE
HDFS Plug-In
HDFS Plug-In
CO-RESIDENT HADOOP ANALYTICS
ANALYTICS WORKSTATION
![Page 36: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/36.jpg)
36 Andreas Koll
Red Hat Storage Server Hadoop connector
2 additional Server
1 additional Server
0 additional Server
FOR TEST ONLY
![Page 37: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/37.jpg)
37 Andreas Koll
Provisioning and life cycle management
Storage
MANAGEMENT TOOLS & FRAMEWORK
Red Hat Storage Server Console
Red HatSatellite
Virtualization
RHEVManager
Simplified and unified storage managementSingle management console for converged storage and compute
Console storage operations● Intuitive user interface
● Volume management
● On-premise and public cloud
● Rest-based API
Provisionable with Satellite● Life cycle management
● Familiar Red Hat Enterprise Linux tools
Virtualization and storage● Shared management with RHEV-M
![Page 38: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/38.jpg)
38 Andreas Koll
Comprehensive, Integrated MonitoringSNMP, extensible Nagios framework - integrates with existing tooling
● Monitoring logical entities and physical resources
● Alerts, reports, trending and capacity planning graphs
● Non-disruptive upgrades
● Works with Red Hat Storage Console
● Can be set up to run in standalone mode
![Page 39: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/39.jpg)
39 Andreas Koll
Red Hat Storage Server subscription options
Red Hat Storage Server is available in 1 and 3 year subscriptions
● Red Hat Storage Server for On-Premise– Designed as a software appliance for deployment within your datacenter
● Red Hat Storage Server Module for On-Premise– Designed for deployment on existing Red Hat Enterprise Linux installs
● Red Hat Storage Server for Public Cloud– Designed for use in the public cloud (amazon currently supported)
● Red Hat Storage Server for Hybrid Cloud– Combines subscriptions for on-premise and public-cloud
![Page 40: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/40.jpg)
40 Andreas Koll
What is in the Red Hat Storage Server subscription?
Red Hat Storage Server for On-Premise is an all-inclusive .iso● Red Hat Enterprise Linux tuned for Red Hat Storage● GlusterFS, XFS file system● Support for all standard protocols● HDFS plugin for Hadoop workloads● Red Hat Storage console
Red Hat Storage Server Module for On-Premise is a collection of RPMs● Deployable on existing Red Hat Enterprise Linux Installations● GlusterFS, XFS file system● Support for all standard protocols● HDFS plugin for Hadoop workloads● Red Hat Storage console
Red Hat Storage Server for Public Cloud is an all-inclusive Amazon AMI● Red Hat Enterprise Linux tuned for Red Hat Storage● GlusterFS, XFS file system● Support for all standard protocols● HDFS plugin for Hadoop workloads ● Does not include Red Hat Storage console
![Page 41: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/41.jpg)
41 Andreas Koll
Hardware (I)
Generic Requirements
● Must be in the Red Hat Hardware Compatibility List for Storage for Red Hat Enterprise Linux 6.0 and newer
● 2-socket (with 4-core, 6-core, or 8-core) servers are recommended
● Reliable backplane RAID controller shipped by server vendors or from OEM manufacturers
● RAID 6 and RAID 1+0 Support in hardware RAID controller (must be flash-backed or battery-backed)
● 1X 50 GB SAS disks for RHSS installation if a separate partition is created for /var (otherwise 200GB)
● 1 X 10 GigE NIC for data traffic is recommended. It's advisable to use NIC bonding with 2 X 10 GigE for increasing throughput and resiliency. 1 GigE NIC may also be used
● Redundant power supply
● Out of band management card to manage and monitor RHS nodes even when the server is down.The same interface may be used for data and management traffic, but it's recommended to have separate networks for data and management.
![Page 42: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/42.jpg)
42 Andreas Koll
Hardware (II)
● High Performance
- 2u/24 (JBODs may be attached)- 15000 RPM 900GB drives(2.5" inch SAS) OR Solid state disks.- Minimum RAM 48 GB
● General Purpose File Serving use-case
2u/12 (JBODs may be attached)7200 or 10000 RPM, up to 6 TB drives (3.5" SAS or SATA)Minimum RAM 32 GB
● Archival use-case
● - 4u/60 (JBODs may be attached)- 7200 or 10000 RPM, up to 6 TB drives (3.5" SAS or SATA)- Minimum RAM 16 GB
![Page 43: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/43.jpg)
44 Andreas Koll
2 Produkte.....
Red Hat Gluster Storage Red Hat Ceph Storage
44 Andreas Koll
![Page 44: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/44.jpg)
45 Andreas Koll
History of CEPH
RHEL-OSP & RHEV Support FEB 2014
MAY 2012Launch of Inktank
OpenStack Integration
2011
2010Mainline Linux Kernel
Open Source 2006
2004 Project Starts at UCSC
2012CloudStack Integration
OCT 2013Inktank Ceph Enterprise Launch
Xen Integration 2013
APR 2014Aquisition by Redhat
![Page 45: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/45.jpg)
TRADITIONAL STORAGE VS. CEPH
Single Purpose
TRADITIONALENTERPRISE STORAGE
Multi-Purpose, Unified
Hardware Distributed Software
Single Vendor Lock-in Open
Hard Scale Limit Exabyte Scale
![Page 46: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/46.jpg)
47 Andreas Koll
CEPH unified storage
FILE SYSTEM
BLOCK STORAGE
OBJECT STORAGE
Keystone
Geo-Replication
Native API
Multi-tenant
S3 & Swift
OpenStack
Linux Kernel
iSCSI
Clones
Snapshots
CIFS/NFS
HDFS
Distributed Metadata
Linux Kernel
POSIX
![Page 47: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/47.jpg)
48 Andreas Koll
CEPH : build for exabyte scale
RADOS
Reliable, Autonomous, Distributed Object Store
Self-healing, self-managing storage nodes
LIBRADOSa library allowingdirect access to RADOSC, C++, JavaPython, Ruby, PHP
bucket basedREST gatewaycompatible with
Swift & S3
fully distributedblockdevice with
Linux Kernelclient and qemu
/KVM driver
POSIX compliantFilesystem with
Linux kernel clientand support
for Fuse
RADOSGW RBD CEPHfs
App Web Host/VM Client
![Page 48: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/48.jpg)
49 Andreas Koll
CEPH : OSD‘s (Object Storage Daemons)
RADOS
Reliable, Autonomous, Distributed Object Store
Self-healing, self-managing storage nodes
FS
OSD
Disk
Disk
FS
OSD
Disk
FS
OSD
Disk
FS
OSD
Disk
FS
OSD
Disk
FS
OSD
Disk
xfs
Softwareagent
![Page 49: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/49.jpg)
50 Andreas Koll
CEPH : OSD‘s & MonitorsDisk
OSD
M
- 10 - 10.00 OSD‘s possible- one / disk, RAID group, SSD- serves stored objects to clients- replication / recovery tasks
- maintains cluster membership & state of OSD‘s- they vote for existing resources- they don‘t serve objects to clients !- provides concensus for distributed decision making- few (3-5) per cluster
![Page 50: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/50.jpg)
51 Andreas Koll
CEPH : OSD‘s & MonitorsDisk
M
M
M
App
![Page 51: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/51.jpg)
52 Andreas Koll
CEPH : Application to Rados Disk
M
M
M
App
Librados
socket connection (fast)
Direct acces to storage nodesfor applications
![Page 52: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/52.jpg)
53 Andreas Koll
CEPH : REST to Rados Disk
M
M
M
App
Librados
Radosgw
socket connection (fast)
REST (Swift / S3)
App App App App App
Loadbalancer Loadbalancer Loadbalancer
Radosgw Radosgw
![Page 53: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/53.jpg)
54 Andreas Koll
CEPH : Multisite Object StoreDisk
M
MM
Librados
Radosgw
Appserver
M
MM
Librados
Radosgw
Appserver
![Page 54: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/54.jpg)
55 Andreas Koll
CEPH : RBD (RADOS Block Device) to Rados Disk
M
M
M
Hypervisor
Librados
RBD
VM
![Page 55: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/55.jpg)
56 Andreas Koll
CEPH : Migration of VMDisk
M
M
M
Hypervisor
Librados
RBD
VM
Hypervisor
Librados
RBD
VM
![Page 56: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/56.jpg)
57 Andreas Koll
CEPH : RBD Device on HostDisk
M
M
M
Host
Librados
Kernel Module RBD
![Page 57: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/57.jpg)
58 Andreas Koll
CEPH integration with OPENSTACK
OPEN STACK
KEYSTONE API
SWIFT API CINDER API GLANCE API
NOVAAPI
CEPH STORAGE CLUSTER(RADOS)
CEPH OBJECT GATEWAY(RGW)
CEPH BLOCK DEVICE(RBD)
HYPERVISOR (Qemu/KVM)
![Page 58: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/58.jpg)
59 Andreas Koll
Usecase : Webhoster
• 15 years creating and deploying services
• Over 340,000 entrepreneur and developer customers
• Open source obsessed› Hosting over 500,000 WordPress sites
› Contributing Ceph, Ceilometer, Akanda
› OpenStack innovator & contributor
![Page 59: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/59.jpg)
60 Andreas Koll
Usecase : Webhoster
36TB PERMACHINE
MM
x12
x12
x12
x12
x12
LOAD BALANCER
LOAD BALANCER MM
MMLOAD BALANCER
LOAD BALANCER
MMradosgw ceph-mon ceph-osd
![Page 60: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/60.jpg)
61 Andreas Koll
Usecase : Webhoster
M
STORAGE NODE
RADOS GATEWAY
RADOS GATEWAY
STORAGE NODE
STORAGE NODE
STORAGE NODE
STORAGE NODE
STORAGE NODE
x90
x4
MANAGEMENT NODE x3
LOAD BALANCER x2STORAGE NODESTORAGE NODE
Dell PowerEdge R5156 core AMD CPU, 32GB RAM2x 300GB SAS drives (OS)12x 3TB SATA drives2x 10GbE, 1x 1GbE, IPMI
Dell PowerEdge R4152x 1TB SATA1x 10GbE
MANAGEMENT NODEMANAGEMENT NODE
![Page 61: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/61.jpg)
62 Andreas Koll
Usecase : Webhoster
x 12 =
3TB 36TB
36TB
x 90 =
~3PB
3TB per OSD12 OSDs per node36TB per node
x=
36TB per node90 nodes~3PB total capacity
x=
/ 3 =~3PB total capacity3 replicas per object~1PB usable capacity
/=
~3PB ~1PB
![Page 62: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/62.jpg)
Andreas Koll63 Andreas Koll63 Andreas Koll63
RED HAT STORAGEFUTURE WORKLOADS
![Page 63: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/63.jpg)
64 Andreas Koll
USE CASES: TODAY AND FUTURE
ANALYTICS
CURRENT USE CASES
Big Data analytics● Storage plug-in for
Hortonworks Data Platform
TARGET USE CASES
Big Data analytics● Persistent back-end for
Spark
Machine data analytics● Online cold storage for
IT operations data with Splunk
Machine data analytics● Storage for ELK, Solr
![Page 64: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/64.jpg)
65 Andreas Koll
USE CASES: TODAY AND FUTURE
OPENSTACK
CURRENT USE CASES
Virtual machine storage● Virtual machine volume
storage with Cinder, Nova and Glance
TARGET USE CASES
Database storage● Storage for relational
databases with Trove
Object storage fortenant applications● Swift-compatible storage
for cloud applications
Storage back-end for Manila● Shared file system-as-a-
service for tenants
![Page 65: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/65.jpg)
66 Andreas Koll
USE CASES: TODAY AND FUTURE
ENTERPRISESHARING
CURRENT USE CASES
Scale out file store● Storage for active archives,
media streaming, content repositories, VM images, and general-purpose file shares
TARGET USE CASES
Compliant archives● Scalable, cost-effective
storage for compliance and regulatory needs
Enterprise filesync and share● Storage for Dropbox-style
enterprise shared folders
File services for containers● File storage services for
containers and pods
![Page 66: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/66.jpg)
67 Andreas Koll
USE CASES: TODAY AND FUTURE
CURRENT USE CASES
S3-based objectstorage for apps● Cost-effective, S3-
compatible, on-premise object store
TARGET USE CASES
Enterprise syncand share● Storage for shared folders
(object backend)
CLOUDSTORAGE
![Page 67: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/67.jpg)
68 Andreas Koll
USE CASES: TODAY AND FUTURE
CURRENT USE CASES
Conventional virtualization storage● Integrated storage for Red
Hat Enterprise Virtualization● (with separate compute and
storage clusters)
TARGET USE CASES
Hyper-converged architectures● Hyper-converged
architectures
ENTERPRISE VIRTUALIZATION
![Page 68: Andreas Koll Solution Architect Storage akoll@redhat€¦ · 11 Andreas Koll Red Hat Gluster Storage synchronous replication with self-healing for server failure asynchronous geo-replication](https://reader033.vdocuments.mx/reader033/viewer/2022053002/5f06807d7e708231d4184ee0/html5/thumbnails/68.jpg)
THANK YOU