GlusterFs Architecture & Roadmap - LinuxCon EU 2013
Download as ODP, PDF2 likes1,762 views
GlusterFS is a scale-out distributed file system that aggregates storage over a network to provide a single unified namespace. It has a modular architecture and runs on commodity hardware without external metadata servers. Future directions for GlusterFS include distributed geo-replication, file snapshots, and erasure coding support. Challenges include improving scalability, supporting hard links and renames, reducing monitoring overhead, and lowering costs.
Ad
More Related Content
What's hot (20)
Viewers also liked (20)
Ad
Similar to GlusterFs Architecture & Roadmap - LinuxCon EU 2013 (20)
Ad
More from Gluster.org (20)
Recently uploaded (20)
GlusterFs Architecture & Roadmap - LinuxCon EU 2013
- 1. GlusterFS – Architecture & Roadmap Vijay Bellur GlusterFS co-maintainer http://twitter.com/vbellur
- 2. 05/17/16 Agenda ● What is GlusterFS? ● Architecture ● Integration ● Use Cases ● Future Directions ● Challenges ● Q&A
- 3. 05/17/16 What is GlusterFS? ● A general purpose scale-out distributed file system. ● Aggregates storage exports over network interconnect to provide a single unified namespace. ● Filesystem is stackable and completely in userspace. ● Layered on disk file systems that support extended attributes.
- 4. 05/17/16 Typical GlusterFS Deployment Global namespace Scale-out storage building blocks Supports thousands of clients Access using GlusterFS native, NFS, SMB and HTTP protocols Linear performance scaling
- 5. 05/17/16 GlusterFS Architecture – Foundations ● Software only, runs on commodity hardware ● No external metadata servers ● Scale-out with Elasticity ● Extensible and modular ● Deployment agnostic ● Unified access ● Largely POSIX compliant
- 7. 05/17/16 GlusterFS concepts – Trusted Storage Pool ● Trusted Storage Pool (cluster) is a collection of storage servers. ● Trusted Storage Pool is formed by invitation – “probe” a new member from the cluster and not vice versa. ● Logical partition for all data and management operations. ● Membership information used for determining quorum. ● Members can be dynamically added and removed from the pool.
- 8. 05/17/16 GlusterFS concepts – Trusted Storage Pool Node2 Probe Probe accepted Node 1 and Node 2 are peers in a trusted storage pool Node2Node1 Node1
- 9. 05/17/16 GlusterFS concepts – Trusted Storage Pool Node1 Node2 Node3Node2Node1 Trusted Storage Pool Node3Node2Node1 Detach
- 10. 05/17/16 A brick is the combination of a node and an export directory – for e.g. hostname:/dir Each brick inherits limits of the underlying filesystem No limit on the number bricks per node Ideally, each brick in a cluster should be of the same size /export3 /export3 /export3 Storage Node /export1 Storage Node /export2 /export1 /export2 /export4 /export5 Storage Node /export1 /export2 3 bricks 5 bricks 3 bricks GlusterFS concepts - Bricks
- 11. 05/17/16 GlusterFS concepts - Volumes ● A volume is a logical collection of bricks. ● Volume is identified by an administrator provided name. ● Volume is a mountable entity and the volume name is provided at the time of mounting. – mount -t glusterfs server1:/<volname> /my/mnt/point ● Bricks from the same node can be part of different volumes
- 12. 05/17/16 GlusterFS concepts - Volumes Node2Node1 Node3 /export/brick1 /export/brick2 /export/brick1 /export/brick2 /export/brick1 /export/brick2 music Videos
- 13. 05/17/16 Volume Types ➢ Type of a volume is specified at the time of volume creation ➢ Volume type determines how and where data is placed ➢ Following volume types are supported in glusterfs: a) Distribute b) Stripe c) Replication d) Distributed Replicate e) Striped Replicate f) Distributed Striped Replicate
- 14. 05/17/16 Distributed Volume ➢ Distributes files across various bricks of the volume. ➢ Directories are present on all bricks of the volume. ➢ Single brick failure will result in loss of data availability. ➢ Removes the need for an external meta data server.
- 15. 05/17/16 How does a distributed volume work? ➢ Uses Davies-Meyer hash algorithm. ➢ A 32-bit hash space is divided into N ranges for N bricks ➢ At the time of directory creation, a range is assigned to each directory. ➢ During a file creation or retrieval, hash is computed on the file name. This hash value is used to locate or place the file. ➢ Different directories in the same brick end up with different hash ranges.
- 16. 05/17/16 How does a distributed volume work?
- 17. 05/17/16 How does a distributed volume work?
- 18. 05/17/16 How does a distributed volume work?
- 19. 05/17/16 Replicated Volume ● Synchronous replication of all directory and file updates. ● Provides high availability of data when node failures occur. ● Transaction driven for ensuring consistency. ● Changelogs maintained for re-conciliation. ● Any number of replicas can be configured.
- 20. 05/17/16 How does a replicated volume work?
- 21. 05/17/16 How does a replicated volume work?
- 22. 05/17/16 Distributed Replicated Volume ● Distribute files across replicated bricks ● Number of bricks must be a multiple of the replica count ● Ordering of bricks in volume definition matters ● Scaling and high availability ● Reads get load balanced. ● Most preferred model of deployment currently.
- 24. 05/17/16 Striped Volume ● Files are striped into chunks and placed in various bricks. ● Recommended only when very large files greater than the size of the disks are present. ● Chunks are files with holes – this helps in maintaining offset consistency. ● A brick failure can result in data loss. Redundancy with replication is highly recommended (striped replicated volumes).
- 25. 05/17/16 Elastic Volume Management Application transparent operations that can be performed in the storage layer. ● Addition of Bricks to a volume ● Remove brick from a volume ● Rebalance data spread within a volume ● Replace a brick in a volume ● Performance / Functionality tuning
- 27. 05/17/16 Access Mechanisms Gluster volumes can be accessed via the following mechanisms: – FUSE based Native protocol – NFSv3 – SMB – libgfapi – ReST/HTTP – HDFS
- 28. 05/17/16 FUSE based native access
- 30. 05/17/16 libgfapi ● Exposes APIs for accessing Gluster volumes. ● Reduces context switches. ● qemu, samba, NFS Ganesha integrated with libgfapi. ● Both sync and async interfaces available. ● Emerging bindings for various languages.
- 31. 05/17/16 libgfapi v/s FUSE – FUSE access
- 32. 05/17/16 libgfapi v/s FUSE – libgfapi access
- 34. 05/17/16 ReST - G4S Client Proxy Account Container Object HTTP Request (REST) Directory Volume FileClient NFS or GlusterFS Mount Unified File and object view. Entity mapping between file and object building blocks
- 37. 05/17/16 Translators in GlusterFS ● Building blocks for a GlusterFS process. ● Based on Translators in GNU HURD. ● Each translator is a functional unit. ● Translators can be stacked together for achieving desired functionality. ● Translators are deployment agnostic – can be loaded in either the client or server stacks.
- 40. 05/17/16 Ecosystem Integration ● Currently integrated with various ecosystems: ● OpenStack ● Samba ● Ganesha ● oVirt ● qemu ● Hadoop ● pcp ● Proxmox ● uWSGI
- 41. 05/17/16 OpenStack Havana and GlusterFS – Current Integration Glance Images Nova Nodes Swift Objects Cinder Data Glance Data Swift Data Swift API Storage Server Storage Server Storage Server… KVM KVM KVM … ● Separate Compute and Storage Pools ● GlusterFS directly provides Swift object service ● Integration with Keystone ● GeoReplication for multi- site support ● Swift data also available via other protocols ● Supports non-OpenStack use in addition to OpenStack use Logical View Physical View libgfapi
- 42. 05/17/16 OpenStack and GlusterFS – Future Integration Glance Images Nova Swift Objects Cinder Data Glance Data Swift Data Swift API Storage Storage Storage … KVM KVM KVM … libgfapi Glance Data Manila Data Savanna Data FileBlock Object
- 43. 05/17/16 GlusterFS & oVirt ● Trusted Storage Pool and Gluster Volume management - oVirt 3.1 ● FUSE based posixFS support for VM image storage - oVirt 3.1 ● libgfapi based Gluster native storage domain - oVirt 3.3 ● Manage converged virtualization and storage clusters in oVirt ● ReST APIs & SDK for GlusterFS management.
- 45. 05/17/16 Use Cases - current ● Unstructured data storage ● Archival ● Disaster Recovery ● Virtual Machine Image Store ● Cloud Storage for Service Providers ● Content Cloud ● Big Data ● Semi-structured & Structured data
- 47. 05/17/16 New Features in GlusterFS 3.5 ● Distributed geo-replication ● File snapshots ● Compression translator ● Multi-brick Block Device volumes ● Readdir ahead translator ● Quota Scalability
- 48. 05/17/16 Beta Features in GlusterFS 3.5 ● Disperse translator for Erasure Coding ● Encryption at rest ● Support for bricks on Btrfs ● libgfapi support for NFS Ganesha (NFS v4)
- 49. 05/17/16 Geo-replication in 3.5 ● Before 3.5 ➢ Merkle tree based optimal volume crawling ➢ Single driver on the master ➢ SPOF ● In 3.5 ➢ Based on changelog ➢ One driver per replica set on the master ➢ No SPOF
- 50. 05/17/16 Quota in 3.5 ● Before 3.5 – Client side enforcement – Configuration in volume files would block scalability – GFID accesses could cause incorrect accounting – Only hard quota supported ● In 3.5 – Server side enforcement – Better configuration management for scalability. – GFID to path conversion enables correct accounting. – Both hard and soft quotas supported
- 51. 05/17/16 Prominent Features beyond GlusterFS 3.5 ● Volume snapshots ● New Style Replication ● pNFS access with NFS Ganesha ● Data tiering / HSM ● Multi master geo-replication ● Support Btrfs features ● Caching improvements ● libgfchangelog ● and more...
- 52. 05/17/16 Challenges ● Scalability – 1024 nodes, 72 brontobytes? ● Hard links ● Rename ● Monolithic tools ● Monitoring ● Reduce Capex and Opex
- 53. 05/17/16 Resources Mailing lists: [email protected] [email protected] IRC: #gluster and #gluster-dev on freenode Links: http://www.gluster.org http://hekafs.org http://forge.gluster.org http://www.gluster.org/community/documentation/index.php/Arch
- 54. Thank you! Vijay Bellur vbellur at redhat dot com