Red Hat Ceph Storage 3: Featuring CephFS and iSCSI support and containerized storage daemons

By Douglas Fuller, Red Hat Ceph Storage Engineering

If you missed last week’s huge announcement about Red Hat Ceph Storage 3, you can find details here. To quickly get you up to speed, though, the big news in this release is around enabling a large variety of storage needs in OpenStack, easing migration from legacy storage platforms, and deploying enterprise storage in Linux containers.

CephFS is here!

One of the highlights of the Red Hat Ceph Storage 3 announcement was production support for CephFS. This delivers a POSIX-compliant shared file system layered on top of massively scalable object storage. Client support is available in the Red Hat Enterprise Linux 7.4 kernel and via FUSE. CephFS leverages Ceph’s RADOS object store for data scalability as well as a natively clustered metadata server (MDS) for metadata scalability, high availability, and performance.

One cluster to do it all

Red Hat Ceph Storage uses the CRUSH structured data distribution scheme, enabling users to deploy a highly scalable and reliable file system using industry-standard, commodity hardware. Expensive, custom-engineered RAID controllers are no longer necessary. Expanding a CephFS deployment is as easy as expanding a Ceph cluster—CRUSH smoothly manages cluster changes, including expansions with new or different hardware.

Have a hybrid storage cluster with SSDs, HDDs, and NVMe devices? CRUSH can divide your storage workload across any and all devices for maximum performance where you need it and maximum capacity at commodity cost where you don’t. This allows disparate workloads—such as scratch, home, or archive data—to coexist in the same cluster using different or overlapping hardware as needed.

In addition, CephFS’s MDS may be dynamically provisioned and resized online to maximize performance and scalability. For metadata-intensive workloads, the Ceph MDS cluster can repartition its workload, either statically or dynamically, online in response to demand. It’s also fault-tolerant by design, with no need for passive standby or expensive and complex “Shoot the Other Node in the Head” (STONITH) configurations to maintain constant availability.

Take the “cluster” out of cluster management

Red Hat Ceph Storage 3 deploys with Red Hat Ansible Automation, integrating smoothly into existing cluster management environments. Now you can deploy and manage compute and storage both using Ansible playbooks.

New in Red Hat Ceph Storage 3 is a REST API for cluster data and management tools. Monitoring tools are available out of the box to provide detailed health and performance data across your Ceph cluster.

A million uses and counting

Red Hat Ceph Storage offers great flexibility to customers. It can be deployed across a wide variety of storage applications, allowing enterprises to manage one unified system supporting block, file, and object interfaces. With the added flexibility of iSCSI support, users from heterogeneous environments—such as VMware and Windows—can leverage the power of the storage platform.

This flexibility is extremely attractive to organizations such as academic research institutions, many of which are participating in the SuperComputing17 conference in Denver this week. Their IT departments have the onerous task of supporting complicated workflows and yet have to work with shoestring budgets in many cases.

To learn more, check out this additional blog post, and join us at the Red Hat SC17 booth (1763) for presentations, swag, and more.

Gartner pegs Red Hat as storage visionary. Two years in a row.

Red Hat Storage strengthens position relative to key competitors

It’s finally here! The Gartner 2017 Magic Quadrant for Distributed File Systems and Object Storage.

We’re extremely excited to announce that Gartner has once again positioned Red Hat in the “Visionaries” quadrant. More important, Red Hat is the furthest to the right and top in the visionary quadrant since last year’s Magic Quadrant. Gartner’s Magic Quadrant judges vendors on completeness of vision and ability to execute.

We are humbled and excited with this new development. It corroborates key investments and strategic product decisions taken over the years by our leadership to deliver tangible and substantial value to customers. The relative movement of Red Hat vis-à-vis established storage vendors tells a compelling story about the rapidly changing enterprise storage landscape.

Some of the key highlights from the past year include:

  • Continued leadership with container-native storage to enable a unified storage platform for cloud-native applications and container infrastructure.
  • Customer traction across geographies and industry verticals that brings to bear Red Hat’s vision of storage for the open hybrid cloud.
  • Strong leadership in upstream open source communities for private cloud infrastructure and Infrastructure as a Service (IaaS), as well as enterprise adoption of highly elastic object storage.
  • Breakthrough innovation around open source hyperconverged infrastructure for remote-office/branch-office and IoT use cases.

You can download a complimentary copy of the Gartner 2017 Magic Quadrant for Distributed File Systems and Object Storage here.

Container-native storage for the OpenShift masses

By Daniel Messer, Red Hat Storage

 

Red Hat Container-Native Storage 3.6, released today, reaches a new level of storage capabilities on the OpenShift Container Platform. Container-native storage can now be used for all the key infrastructure pieces of OpenShift: the registry, logging, and metrics services. The latter two services come courtesy of the new block storage implementation. Object storage is now also available directly to developers in the form of the well-known S3 API. Administrators will enjoy a more robust cns-deploy utility, support for online volume expansion, and more choice in deployment topologies in the OpenShift Advanced Installer. Last, but just as important, it now supports more concurrent workloads serving over 1,000 persistent volumes with just 3 nodes.

________________________________________

You know you must be doing something right when some of your users are looking to use your technology in different ways than expected. Initially, the idea of running GlusterFS alongside Kubernetes and OpenShift promised the ability to use a distributed storage system with a framework for distributed applications. They goes nicely together because both approaches are entirely based on scale-out software, hence independent of the underlying platform, and they are driven by a declarative API-driven design. On the GlusterFS side, that API is available in the form of an additional software daemon, called heketi. Things soon took a new direction when the first experiments of running the GlusterFS/heketi combination as an OpenShift workload were conducted.

A lot of engineering cycles later, the idea of hacking GlusterFS onto OpenShift has emerged to a fully supported product offering: container-native storage. Today, we are happy to announce container-native storage 3.6.

For the impatient: In essence, we have taken container-native storage from being an optional supplement in OpenShift to being a storage solution that now serves file, block, and object storage to applications on top of OpenShift and to the entire OpenShift internal infrastructure, as well.

For the curious reader, let’s go see how we did that….

Increase density

The first thing we had to do was ensure that container-native storage was a robust, scalable, long-term solution for the different possible OpenShift cluster sizes. When we launched container-native storage with OpenShift 3.2 last summer, the container images were based on Red Hat Gluster Storage 3.1.3 and, on average, each brick process on a GlusterFS host/pod consumed about 300 MB of RAM.

That may not sound like much, but you have to be aware that every PersistentVolume served by container-native storage results in a GlusterFS volume being created. Bricks are local directories on GlusterFS pods that make up volumes. The consistency of volumes across all its bricks (by default, 3 in container-native storage) is handled by the glusterfsd process, which is what consumes the memory.

In older releases of Red Hat Gluster Storage, there was one such process per brick on each host. It’s easy to see that with potentially hundreds of application pods in OpenShift requiring their own PersistentVolumes, the resulting number of brick processes in each GlusterFS pod will easily consume gigabytes of RAM and would create a significant effort to coordinate in each pod.

That many processes in a pod are an anti-pattern for Kubernetes and, even if we would have broken out those in separate containers, the memory overhead would still be huge.

Fortunately, Red Hat Gluster Storage 3.3 came to the rescue. Released just a little over 2 weeks ago, it introduced a new feature called brick-multiplexing. It’s easier to depict how this feature changes the structure of a GlusterFS pod in a diagram than a lengthy explanation:

With brick-multiplexing, only one glusterfsd process is governing the bricks such that the amount of memory consumption of GlusterFS pods is drastically reduced and the scalability is significantly improved.

By introducing brick-multiplexing in version 3.6, we are able to support over 1,000 PersistentVolumes in a single container-native storage cluster. The amount of memory consumed increases linearly, so that 32GB of RAM are only needed at the high end of that. The rule of thumb is roughly 30-35 MB RAM per volume on each of the participating GlusterFS pods.

Container-native storage can probably support an even greater number of volumes, and we hope to confirm that soon. Until then, you always have the option to either run more GlusterFS pods in your OpenShift cluster or deploy a second container-native storage cluster, governed by the same Heketi API service.

Optimized storage for logging/metrics

File storage is what containers on OpenShift (and in general) deal with today. It’s a ubiquitous, well-understood concept. There are also proposals for native access to block devices in pods, but they are still in design or planning phases.

That is—at least for now—storage (including block) in Kubernetes and OpenShift always ends up being a mounted file system on the host running the pod, which is then bind-mounted to the target container’s file system namespace. Block storage provisioners in OpenShift eventually format the device with XFS too, before handing it over to the container.

GlusterFS is a distributed, networked file system which, in contrast to local filesystems like XFS, allows shared access from multiple hosts and stores the data in the backend distributed across multiple nodes. This big advantage does not come without cost, however: Some type of operations that are fast and cheap on a local file system are quite expensive in a distributed file system.

For some workloads (e.g., OpenShift Logging and Metrics), this can be a show-stopper. To properly support those, we designed something that might seem counter-intuitive at first: gluster-block. Take a look at the implementation scheme below:

Yes, you see that right: We are using TCM (the Linux kernel’s iSCSI stack, also called LIO) managed by targetcli to create iSCSI LUNs from files on a GlusterFS volume and present those as block devices to pods. The TCM stack allows local storage of a Linux system to be made available on the network via the iSCSI protocol. In our specific case, the local storage is a large raw file on a GlusterFS volume. On the client side, the iSCSI block device will be formatted with XFS and then bind-mounted to the target container’s file system namespace.

But why go through all the trouble? In distributed file systems—and here GlusterFS is no exception—metadata-intensive operations like file create, file open, or extended attribute updates are particularly expensive and slow compared to a local file system. In particular, indexing solutions likes ElasticSearch (part of OpenShift Logging) and scale-out NoSQL databases like Cassandra (part of OpenShift Metrics) generate such workloads.  But also other database software might make heavy use of locking and byte-range locking, which are costly compared to simple read and writes.

In order to qualify OpenShift Metrics and Logging Services to run well on a container-native storage backend, a significant speed up was needed for a lot of special file system operations like these.

You can probably guess what we were thinking: In software, many problems can be solved by adding an additional layer of indirection.

The indirection in accessing data on GlusterFS via iSCSI instead of a normal GlusterFS mount converts otherwise expensive file system operations to a single stream of continuous reads and writes to a single raw file on GlusterFS. The TCM stack delivers this IO stream over the network via iSCSI. On the receiving end, the file in GlusterFS backing the iSCSI LUN is accessed via libgfapi, a userspace library to access files in GlusterFS without the need to mount a volume.

The clients, in our case containers in pods on OpenShift, still write to an XFS file system the iSCSI LUN is formatted with. As a result, simple client-level read and write requests remain virtually as fast as accessing the file directly on GlusterFS, but also all the other file system operations are converted into much faster reads and writes to the file backing the block volume because they are not distributed. From the perspective of GlusterFS, it’s a constant stream of basic read and write requests, which GlusterFS is efficient at. Of course, this comes with a trade-off: gluster-block is not shared storage.

Container-Native Storage version 3.6 now provides backend storage for OpenShift Logging and OpenShift Metrics with gluster-block. For the moment, the use of gluster-block in production is only supported for OpenShift Logging and Metrics services, but use of gluster-block beyond that is under qualification, and support is expected to be extended soon.

The Logging and Metrics services have strict performance and latency requirements and are important for any OpenShift cluster in production. They provide vital information and debugging capabilities for administrators. By design, they are scale-out services, because their storage backend (ElasticSearch for Logging, Cassandra for Metrics) supports a shared-nothing approach. However, in production you do not want additional shards of ElasticSearch and Cassandra run side-by-side with your application pods. That’s why there is a concept of infrastructure nodes in OpenShift that do not run business applications but are dedicated to OpenShift infrastructure components like these. Typically, these kind of servers only have storage locally available, which is limited in capacity and performance. Thus, it might quickly become insufficient to store the logs and metrics of hundred of pods. With container-native storage, you now have a scalable, robust, and long-term storage solution for logging and metrics that utilizes the entire cluster’s storage capacity.

Support a scale-out registry

There is one additional component in OpenShift that’s crucial for operations: the container image registry. This is where all the resulting images from source-to-image builds will be pushed to and where developers can upload their custom images. If it’s unavailable, those operations will fail, and users will be unable to launch new or update existing applications.

The default configuration for the OpenShift registry is to use `emptyDir` storage, that is, a local file system on the container host that depends on the registry pod’s lifetime. In this setup, the registry, of course, cannot be scaled out, updated, or restarted on another host.

Fortunately, as of version 3.5, container-native storage allows for a scale-out registry using shared storage on a PersistentVolume served by GlusterFS. This has several advantages:

  1. No external storage is required, like NFS, which can cause problems with metadata consistency with a busy registry.
  2. There is no dependency on provider storage (e.g., AWS S3 being unavailable in a VMware environment) for shared data access.
  3. The registry can now be scaled out, ideally across all infra nodes.
  4. The registry storage backend can grow dynamically with the platform.

The beauty of this is that it can be installed like this right away. Like we’ve already covered during the announcement of OpenShift Container Platform 3.6 earlier this year, the OpenShift Advanced Installer now supports deploying container-native storage and the registry on container-native storage out of the box. See this video here for details.

All you have to do since OpenShift Container Platform 3.6 is add a few lines to your Ansible inventory file.

To deploy an OpenShift registry backed by container-native storage, first add the following variable definition in the [OSEv3:vars] section:

openshift_hosted_registry_storage_kind=glusterfs

And then add a new host group defining the container-native storage nodes to the inventory, for example:

[glusterfs_registry]
infra-1.lab glusterfs_devices='[ "/dev/sdd" ]'
infra-2.lab glusterfs_devices='[ "/dev/sdd" ]'
infra-3.lab glusterfs_devices='[ "/dev/sdd" ]'

This is enough to tell the OpenShift Advanced Installer that it should create a basic 3-node container-native storage cluster, in this case on the infrastructure nodes, using the supplied devices to create bricks. From this cluster a PersistentVolume will be created and supplied to the registry DeploymentConfig.

That way the registry will be launched with shared storage, provided by container-native storage, and scaled to 3 instances across the infrastructure nodes. You get a highly available and robust registry out of the box with no additional configuration needed.

S3 object storage for applications

In addition to providing block and file storage services, Container-Native Storage 3.6 now provides an S3 object storage interface as a TechPreview. Application developers have a ready-to-use REST API at hand to provide object storage to workloads on OpenShift, just a HTTP PUT or GET request away.

Object storage in Red Hat Container-Native Storage 3.6 provides a simple yet scalable storage layer for distributed applications that were previously tied to specific cloud provider S3 object storage. These application now run with little or no modification on OpenShift.

In this implementation, a gluster-s3 service is deployed as a pod in your OpenShift cluster, and an OpenShift Route is generated for it. The Route’s URL is provided to applications as their S3 endpoint. The service receives the S3 requests and translates those to file system operations on GlusterFS volumes. The S3 buckets and objects are stored as directories and files on that volume, respectively.

For now, this service can be deployed with the cns-deploy utility. There are some new command switches available for this purpose:

cns-deploy topology.json --namespace gluster-storage --log-file=cns-deploy.log --object-account dmesser --object-user dmesser --object-password redhat

The new parameters allow you to specify a name for the S3 account (object-account, an aggregate of multiple S3 buckets, one per CNS cluster), a named user (object-user), and the authentication password for that user in that account (object-password). Once all of these 3 switches are presented, cns-deploy will create the glusterfs-s3 infrastructure.

Support for doing this with the OpenShift Advanced Installer is expected to follow soon. The design foresees exactly one S3 domain/account per CNS cluster, although multiple CNS clusters can be deployed easily.

Improvements for deployment and operations

Besides a whole bunch of new features, we’ve also introduced improvements in usability to make the container-native storage experience better.

In Container-Native Storage 3.6, the cns-deploy tool has been improved in a number of ways. It is now more idempotent, allowing the administrator to run the installer multiple times without having to start from scratch. There will still be error scenarios that may require manual intervention, but it should be much easier to recover from such errors. It will also deploy the required resources to use gluster-block and gluster-s3. Combined with the idempotency improvements, administrators will be able to run cns-deploy to deploy those features into an environment that’s already running container-native storage.

Container Native Storage 3.6 also provides improved integration with container-ready storage. All of our new features will work just as well on container-ready storage as container-native storage. In addition, we have introduced support for a configuration we’re calling Container-Ready Storage without Heketi. heketi is the volume management API service for GlusterFS. In this configuration, container-ready storage runs with the usual Red Hat Gluster Storage nodes outside the OpenShift cluster, but heketi resides as a pod within OpenShift. This has the advantage of making the heketi service highly available rather than residing on a single machine. For new deployments, the cns-deploy can be used to initialize a container-ready storage cluster in this configuration.

Another common scenario that is likely to occur over time, even with the short-lived nature of some workloads, is PersistentVolumes filling to capacity. This can happen when a user under-estimates the required capacity for a workload or the pod simply runs way longer than expected. In any case, heketi now allows for online volume expansion.

To take advantage of this, simply use the heketi-client on the CLI to expand the size of any given volume:

heketi-cli volume expand --volume=0e8a8adc936cd40c2df3698b2f06bba9 --expand-size=2

In the background, heketi changes the GlusterFS volume layout from a 3-way replicated to distributed-replicated. See below for a comparison from GlusterFS perspective.

Before volume expansion:

sh-4.2# gluster vol info vol_0e8a8adc936cd40c2df3698b2f06bba9

Volume Name: vol_0e8a8adc936cd40c2df3698b2f06bba9
Type: Replicate
Volume ID: 841bd097-659b-4b5d-b3ec-56bb8cc51c2f
Status: Started
Snapshot Count: 0
Number of Bricks: 1 x 3 = 3
Transport-type: tcp
Bricks:
Brick1: 10.20.5.232:/var/lib/heketi/mounts/vg_c05319c8a95eaa083adbedb7d43913fa/brick_4bf9ae183dacceccf4bf525186850bdd/brick
Brick2: 10.20.6.239:/var/lib/heketi/mounts/vg_bd7fbf9053d6340771f7b75ce2872339/brick_e1175aaaa8596aedc18bf8c56b42fe8d/brick
Brick3: 10.20.4.184:/var/lib/heketi/mounts/vg_0797a1d458309eec3e5e818a9b87f6c6/brick_2b5255cc2c0297e4e34eb6f1b4319fb9/brick
Options Reconfigured:
transport.address-family: inet
nfs.disable: on
cluster.brick-multiplex: on

After volume expansion:

sh-4.2# gluster vol info vol_0e8a8adc936cd40c2df3698b2f06bba9

Volume Name: vol_0e8a8adc936cd40c2df3698b2f06bba9 
Type: Distributed-Replicate 
Volume ID: 841bd097-659b-4b5d-b3ec-56bb8cc51c2f 
Status: Started 
Snapshot Count: 0 
Number of Bricks: 2 x 3 = 6 
Transport-type: tcp 
Bricks: 
Brick1: 10.20.5.232:/var/lib/heketi/mounts/vg_c05319c8a95eaa083adbedb7d43913fa/brick_4bf9ae183dacceccf4bf525186850bdd/brick 
Brick2: 10.20.6.239:/var/lib/heketi/mounts/vg_bd7fbf9053d6340771f7b75ce2872339/brick_e1175aaaa8596aedc18bf8c56b42fe8d/brick 
Brick3: 10.20.4.184:/var/lib/heketi/mounts/vg_0797a1d458309eec3e5e818a9b87f6c6/brick_2b5255cc2c0297e4e34eb6f1b4319fb9/brick 
Brick4: 10.20.6.239:/var/lib/heketi/mounts/vg_bd7fbf9053d6340771f7b75ce2872339/brick_c48d4ea4b43635f62c464ddf0259d733/brick 
Brick5: 10.20.4.184:/var/lib/heketi/mounts/vg_0797a1d458309eec3e5e818a9b87f6c6/brick_121fbc266c905311d8a8810f221fbdca/brick 
Brick6: 10.20.5.232:/var/lib/heketi/mounts/vg_c05319c8a95eaa083adbedb7d43913fa/brick_5f208c680444b4820f53c923aa079614/brick 
Options Reconfigured: transport.address-family: inet nfs.disable: on cluster.brick-multiplex: on 

Finally, with Container-Native Storage 3.6, we have expanded the amount of technical documentation available. We provide more examples of things both new and pre-existing that you can do with container-native storage, as well as detailed upgrade procedures from a variety of configurations to make sure you can get the latest set of features.

Verdict

The storage play for containers is an exciting space at the moment. There are many options available for customers, and Red Hat container-native storage is unique in the way it runs natively on OpenShift and provides scalable shared file, block, and object storage to business applications and container platform infrastructure.

OpenShift Container Platform 3.6: Streamlined installation and configuration of Red Hat Gluster Storage for containers

By Erin Boyd, Jose Rivera, and Scott Creeley, Red Hat

Did you ever get a new, exciting toy only to have that excitement squashed by the phrase “Batteries not included”?

With the introduction of Red Hat OpenShift Container Platform 3.6, no longer will customers have to wait or jump extra hurdles to get resilient, persistent storage with their new installations. Now they can more easily deploy Red Hat Gluster Storage ready for use by their containerized applications—This is PaaS with batteries included!

With the release of Red Hat OpenShift Container Platform 3.6, users will have the convenience of using a single tool to use Red Hat Gluster Storage as either container-native storage (CNS) or container-ready storage (CRS) alongside the rest of their OpenShift installations. As part of the OpenShift Advanced Installation, users can specify two new storage options: Red Hat Gluster Storage for (1) hosted registry storage or (2) general application storage. To facilitate evaluation of these, an Openshift Container Platform evaluation subscription now includes Red Hat Gluster Storage evaluation binaries and subscriptions.

Following is a sample inventory file that would be used with an OpenShift Container Platform Advanced Installation that deploys two CNS clusters for both hosted registry storage and general application storage.

[OSEv3:children]
 masters
 nodes
 glusterfs_registry
 glusterfs

[OSEv3:vars]
 ansible_ssh_user=root
 openshift_master_default_subdomain=cloudapps.example.com
 openshift_deployment_type=openshift-enterprise
 openshift_hosted_registry_storage_kind=glusterfs
 openshift_disable_check=disk_availability,memory_availability

[nodes]
 master1 node=True storage=True master=True openshift_schedulable=False
 node1 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node2 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node3 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node4 node=True storage=True openshift_schedulable=True
 node5 node=True storage=True openshift_schedulable=True
 node6 node=True storage=True openshift_schedulable=True

[glusterfs_registry]
 node1 glusterfs_devices="[ '/dev/xvdc' ]"
 node2 glusterfs_devices="[ '/dev/xvdc' ]"
 node3 glusterfs_devices="[ '/dev/xvdc' ]"

[glusterfs]
 node4 glusterfs_devices="[ '/dev/xvdc' ]"
 node5 glusterfs_devices="[ '/dev/xvdc' ]"
 node6 glusterfs_devices="[ '/dev/xvdc' ]"

[masters]
 master1 node=True storage=True master=True openshift_schedulable=False

Let’s go over the highlighted portions in detail.

The first section defines the host groups the installation will be using. We’ve defined two new groups: (1) glusterfs_registry and (2) glusterfs. The first specifies a cluster that will host a single volume for use exclusively by a hosted registry. The second specifies a cluster for general application storage and will, by default, come with a Storage Class to enable dynamic provisioning.

[OSEv3:children]
 masters
 nodes
 glusterfs_registry
 glusterfs

In the following section, we indicate that we want the hosted registry to use Red Hat Gluster Storage for its storage needs.

[OSEv3:vars]
 ansible_ssh_user=root
 openshift_master_default_subdomain=cloudapps.example.com
 openshift_deployment_type=openshift-enterprise
 openshift_hosted_registry_storage_kind=glusterfs
 openshift_disable_check=disk_availability,memory_availability

In the [nodes] section, we need to specify all nodes in the OpenShift Container Platform cluster. For our installation, we also need to specify which nodes will run pods for the hosted registry. This is done by specifying “openshift_node_labels=”{‘region’: ‘infra’}”” for each such node. It is recommended to have at least three nodes running your hosted registry.

[nodes]
 master1 node=True storage=True master=True openshift_schedulable=False
 node1 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node2 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node3 node=True storage=True openshift_node_labels="{'region': 'infra'}"
 openshift_schedulable=True
 node4 node=True storage=True openshift_schedulable=True
 node5 node=True storage=True openshift_schedulable=True
 node6 node=True storage=True openshift_schedulable=True

Now we get to our new sections where we specify the nodes that will be used for storage. CNS and CRS require that each cluster have a minimum of three nodes. Multiple clusters can not share a given node. Because we are deploying two clusters, we need to specify six nodes total. It is also required that each node have at least one dedicated, bare storage device (no data or formatting of any kind) for exclusive use by Red Hat Gluster Storage.

Our first new section is [glusterfs_registry]. Here we specify the nodes of the Red Hat Gluster Storage cluster and the storage devices on those nodes that will be used for a hosted registry’s storage. It is not required that these nodes be the same as the ones running the hosted registry.

[glusterfs_registry]
 node1 glusterfs_devices="[ '/dev/xvdc' ]"
 node2 glusterfs_devices="[ '/dev/xvdc' ]"
 node3 glusterfs_devices="[ '/dev/xvdc' ]"

Our second new section, [glusterfs], is used for specifying the Red Hat Gluster Storage cluster and storage devices that will be used for general application storage. These storage devices must also be for exclusive use by Red Hat Gluster Storage. As mentioned, these nodes may not also be part of the cluster used by [glusterfs_registry]. In the case of CNS, it is not required that these nodes be dedicated exclusively to serving storage; CNS pods can coexist with other application pods.

[glusterfs]
 node4 glusterfs_devices="[ '/dev/xvdc' ]"
 node5 glusterfs_devices="[ '/dev/xvdc' ]"
 node6 glusterfs_devices="[ '/dev/xvdc' ]"

Once the installer is complete, the user can see the pre-defined Storage Class by executing:

# oc get storageclasses
 NAME TYPE
 glusterfs-storage kubernetes.io/glusterfs

This Storage Class can be used for applications by specifying a Persistent Volume Claim to dynamically provision the required storage volume:

apiVersion: v1
 kind: PersistentVolumeClaim
 metadata:
 name: mypvc
 namespace: glusterfs
 spec:
 accessModes:
 - ReadWriteOnce
 resources:
 requests:
 storage: 100Gi
 storageClassName: glusterfs-storage

And that’s it, your PaaS solution with built-in storage is ready to go! If you want to tune the installation further, more options are available in the Advanced Installation, and a demo video is available here.

Introducing Red Hat Hyperconverged Infrastructure 1.0

By Steve Bohac, Red Hat Storage

Today we’re proud to announce Red Hat Hyperconverged Infrastructure 1.0! By combining Red Hat virtualization and storage technologies with a stable, proven operating platform, Red Hat Hyperconverged Infrastructure is designed to help enterprises bring datacenter capabilities into locations with limited space, such as branch offices and other remote facilities.

Built on Red Hat Virtualization and Red Hat Gluster Storage, Red Hat Hyperconverged Infrastructure provides simplified planning and procurement, streamlined deployment and management, and a single support stack for virtual compute and virtual storage resources. Red Hat Hyperconverged Infrastructure is an ideal solution for remote/branch office or edge computing needs. Deployment is enabled by Ansible by Red Hat, and Red Hat CloudForms can be used to manage all the Red Hat Hyperconverged Infrastructure installations in your enterprise via a single application.

Customers have been asking us for this type of an integrated solution, so we’re happy to offer this hyperconverged combination in a single SKU to satisfy that request.

Organizations with distributed operations, such as those in the banking, energy, or retail industries, can benefit from offering the same infrastructure services in remote and branch offices as they run in their datacenters. However, remote and branch offices can have unique challenges: less space and power/cooling and fewer (or no) technical staff on site. Organizations in this situation need powerful services, integrated on a single server that allow them to keep their key applications local to the remote site.

Red Hat Hyperconverged Infrastructure addresses these challenges for remote installations. The following figure depicts the benefits that consolidation with a hyperconverged infrastructure provides:

  • Eliminate storage as a discrete tier
  • Easily virtualize business applications, maximizing resource utilization
  • Single budget for compute and storage
  • Single team managing infrastructure
  • Simplified planning and procurement
  • Streamlined deployment and management
  • Single support stack for compute and storage

Removing the storage tier by consolidating compute and storage onto a single server platform/tier offers streamlined deployment and management (enabled by Ansible by Red Hat and Red Hat CloudForms), a single support stack (one vendor to call now instead of two), and simplified planning and procurement (reducing the number of vendors to source from).

For more information on Red Hat Hyperconverged Infrastructure, click here.

For an on-demand webinar discussing Red Hat Hyperconverged Infrastructure in more detail, click here.

Storage can make your digital transformation—or break it

By Ross Turk, Red Hat Storage

The following chart might look familiar, especially if you’ve ever studied patterns of online behavior.

Like all the best charts, this one has a glorious up-and-to-the-right shape. But each year at the end of December, when much of the world goes offline for a few quiet days, there’s a characteristic drop. This chart—from Google Trends—represents a phrase that’s growing in prominence: “digital transformation.”

Digital transformation is everywhere

When I noticed—with distinct déjà vu—the industry using this phrase, I admit I was somewhat taken aback. Many of us live in a world dominated by technology. I can’t remember the last time I paid for fast-food tacos with actual money, but I do know I stopped carrying cash completely when the taco shops began accepting credit cards. Every time I need to mail a letter now, it’s a huge production! I’m just not prepared for that kind of task anymore.

Imagine a world without digital technology…. See?! You can’t do it.

Not all digital transformation is equal

Another case in point. I renewed my driver’s license recently and found myself wondering: Now that the DMV is doing business using modern technology, who’s left to transform? If you live your life in an ivory tower made of wifi and capacitive touch screens—like me—a phrase like “digital transformation” can seem obsolete. It can throw you off the scent a bit. And, indeed, I was missing the point. Sure, even taxi companies embrace digital technologies these days…but are they any good at it? Do they enjoy the same efficiencies as a digitally native service like Uber?

Technology is now serious business

Digital transformation isn’t about using technology—or even offering digital services. It’s about redefining a business in technology terms, putting the modern technology experience first. It’s about businesses coming to terms with the truth: Technology can’t be a hobby for them anymore. They’re going to have a ton of applications and data, and they need to get really good at managing all of it. That means having solid priorities; agility and elasticity are a great place to start.

Modern storage can transform your business

Speaking of great places to start, there’s no better example of the challenges of digital transformation than storage. The amount of data that enterprises need to maintain is growing at a steady clip, and their customers expect all that data to be available instantly. Access patterns change as frequently as customer behaviors. Data is getting bigger, analytics are getting even more sophisticated. The traditional storage appliances that do a lot of the heavy lifting today are convenient, but at petabyte scale they show their inflexibility and limitation.

That’s where Ceph and Gluster come in. They’re flexible, scale-out, software-defined storage technologies built for those who don’t think storage is a hobby.

Learn how storage can make your digital transformation

If you want to learn more about modern approaches to storage—and Red Hat Storage, of course!—join me on June 22 for a 45-minute webinar. Register here.

Going public with Red Hat Ceph Storage 2.3

By Daniel Gilfix, Red Hat Storage

You may not have heard a lot about Red Hat Ceph Storage lately, but that doesn’t mean the product hasn’t been active. News in conjunction with Red Hat OpenStack Platform 11 and 10, technology partners such as Rackspace and Micron, and customer adoption at places like Massachusetts Open Cloud, UKCloud, and CLIMB have reinforced the product’s role as a cornerstone of the Red Hat portfolio. But the advances of the product, itself, have been relatively under wraps, with versions 2.2 and 2.1 carefully monitored by existing fans and loyal software-defined storage blog readers. We don’t expect the announcement of Red Hat Ceph Storage 2.3 to shake mountains with seismic impact, but we do expect it to inspire our user community with the doors the product opens today and what might be possible long term.

Source: Sheri Terris, from June 1, 2007, https://www.flickr.com/photos/crestedcrazy/534647428

Greater versatility

Red Hat Ceph Storage 2.3 takes aim at extending the versatility of object storage so that users can connect more successfully to traditional workloads and link them effectively to modern ones. One way is through our new Network File System (NFS) gateway into the product’s S3-compatible object interface. The gateway facilitates the adoption of Ceph Storage as a target for file-based data without requiring changes in data access protocols or the management of data caching semantics. It means Red Hat Ceph Storage users can access the same data set from both object and file interfaces and gradually transition between them based on business need. It also means they can extend the multi-site capabilities of Red Hat Ceph Storage to enable global clusters and data access with the NFS protocol.

Improved connectivity

Red Hat Ceph Storage 2.3 is laying the groundwork for improved connectivity with analytics engines. By conducting validation tests for compatibility with the Hadoop S3A plugin, Red Hat is extending financial and operational benefits of a scale-out, software-defined object storage platform to analytics workloads and data-driven applications leveraging tools like Hadoop, Hive, and Spark. With analytics data stored in an S3-compatible object store, developers have access to a broader ecosystem of tools and language bindings, no longer forced to use a bulky HDFS client. By concentrating data sets in a common object store, data duplication and data lineage challenges are simplified. Multiple ephemeral instances of elastic analytics clusters can reference a single source of truth.

Deployment flexibility

A final capability targets the highly coveted customer requirement of running Ceph in a containerized format. Red Hat Ceph Storage 2.3 includes a single container image that delivers the same capabilities as in the traditional package format. With a Red Hat Ansible-based deployment tool, users can perform installations, upgrades, and updates with atomicity for reduced complexity, easier management, and faster deployment. This supports customers in areas like telecommunications seeking standardized orchestration and deployment of infrastructure software in containers with Kubernetes by adding a cloud-native storage service to this orchestration methodology.

A new chapter

These new capabilities demonstrate the new heights Red Hat Ceph Storage aims to scale, supported by a company firmly committed to real-world deployment of the future of storage. By combining massive scalability with multi-protocol access to highly available clusters, Red Hat Ceph Storage is moving object storage up the mountain to help unleash its power for modern workloads. For more information, check out this short video:

Red Hat Ceph Storage 2.3 is expected to be available later this month.

Storage for the modern enterprise

In early May, Red Hat put on its annual customer and community conference in Boston—Red Hat Summit—centered around a common theme: the power of the individual. Now more than ever, open source is seen as the most viable option as we enter the next phase of IT delivery, planning, and deployment.

The power of the individual on display

Red Hat has assumed the role of de facto open source leader, driving and nurturing hundreds of communities across the world. One could argue that, at the core, Red Hat isn’t a software company at all. In fact, our best asset is our ability to curate open source communities, bringing to bear the efforts of thousands of contributors, committers, and testers to enterprises in a reliable, secure package that can solve some of the most demanding IT challenges.

The end of planning as we know it

Red Hat CEO, Jim Whitehurst, made a pertinent point in his keynote about the changing face of IT planning. “Planning harder” in an environment full of unknowns is complex and fraught with error. CIOs struggle to balance predictability with the inherent flexibility needed to maintain smooth IT operations.

Building IT infrastructure with open source and industry-standard constructs helps alleviate some of the planning challenges while addressing today’s pain points, much like interchangeable Lego® pieces can be used to build everything from two-story houses to skyscrapers, still withstanding the shock of an earthquake when needed.

Storage for the modern enterprise

With each passing year, it becomes apparent that traditional storage just isn’t going to cut it as enterprises look to create flexible, scalable, and cost-effective IT platforms for cloud-native applications. Simply put, legacy storage systems have failed to keep up with the way customers want to consume storage.

A key piece of the value proposition of software-defined storage is the hardware choice available to customers. For the second year in a row, a Storage Ecosystem Showcase in the partner pavilion of Red Hat Summit featured seven Technology Partners that complete or enhance Red Hat’s software-defined storage offering. Cisco, NGD Systems, Permabit, QCT, Seagate, Storage Made Easy, and Supermicro all demonstrated their wares.

In addition, several other storage partners, such as Dell EMC, Mellanox, and Penguin Computing, chose to sponsor their own booths. The solid upstream ecosystem combines with a growing downstream array of partners to truly differentiate Red Hat Storage.

Learn more

Storage was featured prominently both on the expo floor and in the news from the event. In addition to breakout sessions, Red Hat Storage engineers and consultants held a number of hands-on labs that were very well received. You can access similar self-paced material on the online AWS test drives (Gluster test drive and Ceph test drive) or at an upcoming webinar.

For a review of all the 2017 Red Hat Summit videos, click here. For a video recap with some of the Red Hat Storage team, watch this:

Moving the ball forward for object storage with Red Hat Ceph Storage 2.2

We are happy to announce that Red Hat Ceph Storage 2.2 is now Generally Available (GA). It is based on the open source community version of Ceph Storage, specifically version 10.2.5, the Jewel release stream. Similar to Red Hat Ceph Storage 2 announced last summer, Red Hat Ceph Storage 2.2 has a heavy focus on object storage deployments. In addition to following a new, more predictable release process, Red Hat Ceph Storage 2.2 offers a number of enhancements, including:

Global clusters

The object access method for Red Hat Ceph Storage (aka the RADOS Gateway, or “RGW”) now supports up to three sites for global cluster configurations. This means that customers can deploy an active-active global cluster across three geographically distributed sites with data replication and consistency across all three. Alternatively, the RGW multi-site capability can be used in disaster recovery configurations for data protection against site disaster using an active-passive deployment.

Better security and encryption support

Red Hat Ceph Storage now has native support for the Secure Sockets Layer (SSL) in RGW. This is a good option for small to medium size installations that require data encryption between the object storage client and RGW. Because SSL and encryption can have a performance impact, we still recommend terminating SSL at the load balancer or HA layer for large-scale installations.

S3 API enhancements

Red Hat Ceph Storage supports the S3 object API for easy application mobility across private and public clouds. Red Hat Ceph Storage 2.2 now adds support for the new multipart upload copy-part S3 API.

Swift enhancements

Red Hat Ceph Storage 2.2 also has important developments for customers using the Swift object API that demonstrate our continued focus on object storage deployments, including:

  • Support for Swift object versioning functionality.
  • Full testing and compliance of Red Hat Ceph Storage 2.2 object storage with the Swift API tests in the Tempest test suite from RefStack toolset for OpenStack. For customers, this translates into interoperability between applications and services using the Swift API and Red Hat Ceph Storage.

For a recent example of a customer success with Red Hat Ceph Storage at scale for object storage, check out our recently published success story with the CLIMB project in the U.K.

For general information on object storage features in Red Hat Ceph Storage, read this blog.

For general information on Red Hat Ceph Storage, visit this page.

The value of on-premise object storage

The recent outage of the Amazon Web Services (AWS) S3 offering has likely reminded customers of the old “too many eggs in one basket” adage, sparking reviews of where and how they deploy and manage their storage. Outages of this magnitude illustrate how much data lives in object storage these days.

Object storage has come to be the foundation for web-scale architectures in public or private clouds, as we’ve previously discussed. Its allure is due to its potential to handle massive scale while minimizing complexity and cost. Customers struggling with large-scale data storage deployments are turning to object storage to overcome the limitations that legacy storage systems face at scale (typically degradation of performance).

Object storage allows application developers and users to focus more on their workflow and logic and not worry about handling file storage and file locations. However, the large extent of outages due to the unavailability of AWS S3 shows the dependence that so many businesses have on public-cloud-based object storage today.

Ceph is an object store at its core and was designed for web-scale applications. Running Red Hat Ceph Storage on premises offers customers a more secure, cost-effective, and performant way to manage large amounts of data without the risks that can be associated with public-cloud-based solutions.

The S3 API has become the de facto standard for accessing data in object stores, and organizations now develop and demand applications that “speak” S3. Applications that use the S3 API can more easily migrate between on-premise, private clouds built on Red Hat Ceph Storage and public clouds like AWS just by changing the storage endpoint network address. This enables them to operate on both datasets that are present in private clouds and those stored in the public cloud. A common API for object storage means that applications can move between these two cloud deployment models, managing data consistently wherever they go.

For a recent customer example using Red Hat Ceph Storage at scale for object storage, check out our recently published success story with the CLIMB project in the U.K.

Hopefully, most of you reading this were not stung by the recent outage. Regardless, now is as good a time as any to review your infrastructure to determine if an on-premise object storage approach with Red Hat Ceph Storage makes sense. We think it does!

For more information on Red Hat Ceph Storage for object storage, check out this page.

For more information on Red Hat Ceph Storage for cloud Infrastructures, check out this page.

Cue the drumroll…Red Hat Gluster Storage 3.2 releases today

Faster metadata operations, deeper integration with OpenShift, smaller hardware footprint

It’s here! Red Hat today announced the general availability of Red Hat Gluster Storage 3.2. Please join the live launch event at 12pm EDT today to hear from customers, partners, experts, and the thriving open source GlusterFS community. Can’t make it? No sweat. The content will be available for your viewing pleasure for 90 days.

The best yet. And then some.

This release is a significant milestone in the life of the product. It brings to bear a number of key enhancements but also represents a solid, stable release that is a culmination of much of the innovation over the past 18 months.

Red Hat Gluster Storage is highly scalable, software-defined storage that can be deployed anywhere – from bare metal, to virtual machines, to containers, and across all three major public cloud providers (Microsoft Azure, Google Cloud Platform, and Amazon AWS).

Red Hat Gluster Storage 3.2 is geared toward cloud-native applications and modern workloads that demand a level of agility and scale that traditional storage appliances cannot offer. Read on to find out how this release delivers performance and cost improvements along with a seamless user experience.

Deeper integration with OpenShift Container Platform

Container-native storage for OpenShift Container Platform enables developers to provision and manage storage as easily as they can applications. In addition, improvements in small-file performance in Red Hat Gluster Storage will help container registry workloads – the heart of the container platform itself – to persist Docker images for containerized applications.

In addition to small file performance, there are a number of scale improvements to the management plane that allow for a larger number of persistent volumes (OpenShift PVs) that can be hosted in a single cluster. Also, the container for Red Hat Gluster Storage has been configured to run the sshd service for supporting async geo-replication between remote sites and has been instrumented to hold the SSL certificates for supporting in-flight encryption to be leveraged by container-native storage in upcoming releases.

Storage administrators will also find that better small file performance allows for a better user experience with day-to-day operations, such as directory listings, up to 8x faster. The secret sauce, client-side caching, that facilitates this performance improvement will benefit both Linux (Fuse) and Windows (SMB) users.

The improvement in small file performance will be particularly relevant to use cases with a large number of files under a few MBs in size. A number of use cases, such as hosting Git repositories and electronic design automation, will benefit as well.

Lower hardware footprint and costs

Most companies look to 3-way replication for enterprise-grade data integrity. What if we told you that you could benefit from the same grade of data integrity while lowering your hardware footprint and cost?

This is possible in Red Hat Gluster Storage 3.2 through a new type of volume known as an arbiter volume that does just that – arbitrates between two nodes that may be out of sync. This is particularly useful for remote office-branch office (ROBO) scenarios where remote locations are interested in saving space, power, and cost through hyperconvergence.

Additionally, faster self-healing of sharded storage volumes will benefit customers running hyperconverged configurations, such as Red Hat Virtualization with Red Hat Gluster Storage.

Performance enhancements

Multi-threaded applications can perform faster by parallelizing I/O requests by enabling client-io threads. This feature is activated automatically during heavy loads and dials back during idle state. There are significant performance gains with more number of threads with client io threads enabled.

Our internal test results have shown up to 250% performance increase with the dispersed volumes with client-io-threads enabled. There is a near linear performance gain with the increase in the number of threads.

Native eventing and integration with Nagios

Red Hat Gluster Storage 3.2 features a new native push-based notification framework that can send out asynchronous alerts. As with earlier releases, full integration with the open source Nagios framework is supported.

Learn more. Speak to Red Hat experts and customers.

Find out more about the benefits of the new feature-rich Red Hat Gluster Storage at our launch event happening today at 12pm EDT. Sign up for free, and interact directly with our experts who will be standing by to answer your questions.

Hear from customers like Brinker International who are running containers on Red Hat Gluster Storage in production and have some great results to share. Hear also about great things happening in the GlusterFS community, which includes contributors like Facebook. We hope to see you there.

Cloud native app developers delight: Container storage just got a whole lot easier!

The new Red Hat OpenShift Container Platform offers a rich user experience with dynamic provisioning of storage volumes, automation, and much more

 

By Michael Adam, Engineering Lead, Container Native Storage, and Sayan Saha, Head of Product, Red Hat Gluster Storage

 

Today, Red Hat announced general availability of Red Hat OpenShift Container Platform 3.4, which includes key features such as enhanced multi-tenancy and streamlined deployment for hybrid clouds. In addition, a number of open source storage innovations have been included in this release, which enable easier storage management and provisioning across the lifecycle of containers.

The story so far

Containers were built to be ephemeral and stateless. However, stateful applications running in containers need enterprise-grade persistent storage. Over the past 18 months, Red Hat has delivered a continuum of innovation around persistent storage for containers, leading the charge on both fronts – the open source communities and enterprise products. Red Hat offers container-native storage – durable, distributed, software-defined storage integrated deeply into the Red Hat OpenShift Container Platform, managed by Kubernetes.

cns3-4-1

Rich developer and management experience

In the latest release, Red Hat OpenShift Container Platform 3.4 offers dynamic provisioning of persistent volumes, allowing for a much richer developer experience, addressing annoying delays due to lengthy storage provisioning cycles needed by traditional storage platforms.

Storage administrators can expect to find that easier volume management with dynamic provisioning frees them up for more value-added tasks. Developers building cloud-native apps deployed in containers can benefit from faster storage provisioning and a better user experience.

DevOps managers can relish the automation and integration through a new deployment tool included with the subscription that can deploy container-native storage with push-button simplicity.

Dynamic provisioning for persistent volume claims

Prior to this release, storage administrators and application developers were limited to a static provisioning model where persistent volumes (PVs) of fixed capacity had to be pre-provisioned manually to be consumed by applications running in Kubernetes pods.

Persistent volume claims (PVCs) are used to consume storage resources in Kubernetes like pods that consume compute resources. When new PVCs were received, an attempt was made to match the PVC request with the closest available PV in terms of capacity, and if one was found the claim would be bound to it. This scheme is inefficient.

Consider a situation where 10, 100 GB PVs have been pre-provisioned and made available. A request for 50 GB of storage would be matched to one of the available 100 GB PVs. This is wasteful as storage is over-committed. On the other hand, a request for 150 GB of storage would go unsatisfied as there is no close match, even though there is unused storage capacity.

The new dynamic provisioning feature fixes that issue by automating the provisioning of storage volumes. For instance, a 50 GB PVC request is addressed using a 50 GB PV that is dynamically provisioned for developers requiring zero admin intervention. In other words, users can expect to get exactly what they asked for as long as the underlying storage platform has available capacity.

Note that dynamic provisioning is supported even when Red Hat Gluster Storage serves out storage from a dedicated storage cluster in addition to container-native storage. This demo shows how container-native storage can be dynamically provisioned in OpenShift Container Platform.

cns3-4-2

Dynamic provisioning using storage classes

Dynamic provisioning is enabled by a new feature in OpenShift called storage classes. Storage classes enable storage admins to describe as well as classify their various storage implementations that are available to be used by the OpenShift cluster, and they enable developers to configure specific parameters when requesting storage on demand. Container-native storage can be configured as a storage class, which allows OpenShift developers to dynamically provision storage when submitting claims against the storage class, as seen below.

cns3-4-3

Faster and easier storage deployments using Kubernetes daemon sets

Container-native storage now ships with a deployment tool that will deploy the whole system in an already installed OpenShift cluster. The deployment tool is flexible in that it can easily be used in Ansible playbooks. The administrator only needs to prepare a topology file, a JSON-formatted file describing the nodes and storage devices to be used. Based on that, the deployment of the Gluster storage cluster and the management server as pods in the OpenShift cluster is achieved with the invocation of just a single command. Once deployment is completed, the Gluster storage is ready for both manual and dynamic provisioning with an appropriate storage class. In case of any errors encountered during deployment, the tool supports an abort operation that undoes the failed partial deployment, so that it can be started from scratch. This demo shows the deployment tool in action.

cns3-4-4

GID level security and endpoints

Several features have been added to Red Hat OpenShift Container Platform 3.4 to create a more secure storage environment. The first of these is the addition of system-controlled, preallocated GIDs for the Red Hat Gluster Storage container. This enables the container to run as a non-root user, permitting only allowed users to access the data.

Second, usability with endpoints has been resolved with the deployment of a service and endpoint for each dynamically provisioned volume. This allows PVs to be specific to the requestors namespace without the added steps of manually creating these resources.

The most comprehensive persistent storage for containers

Red Hat continues to be a major contributor to the Docker and Kubernetes communities. In fact, as of today, Red Hat has the second-most contributors in each, second only to Docker and Google, respectively. Much of the innovation happening upstream is focused on solving the persistent storage challenge for stateful applications. Red Hat has contributed a number of volume plugins for a variety of protocols. Learn more about the latest innovations from Red Hat during the virtual event on January 19 or in a webinar with container storage experts on January 24. Learn more at redhat.com/containerstorage.