Upgrading from OneFS 6.5: Ensuring a Smooth Upgrade Process

NOTE: This topic is part of the Uptime Information Hub.

 

OneFS 6.5.x will reach the end of its service life (EOSL) on June 30, 2015. We strongly encourage customers who are running OneFS 6.5.x to upgrade to a supported version of OneFS. After June 30, if you call Support to open a case, you will be directed to upgrade to a supported code family. Support might still triage your issue, but will do so only on a limited basis. No code releases or patches will be offered or available for the 6.5 code family after June 30, 2015. For links to more information about upgrading from OneFS 6.5, visit the OneFS 6.5 Upgrade Info Hub.

 

Upgrading to OneFS Target Code

 

OneFS Target Code is a version of OneFS that has been proven through extensive use in the field to surpass even the high bar set for all OneFS releases, meeting stringent criteria such as a minimum number of runtime hours, the number of clusters that the code is installed on, and clusters that have each licensed OneFS feature enabled. The OneFS 7.1.1 family is the version that has demonstrated the best balance of high adoption rates and low data unavailability/data loss rates. The OneFS 7.2.0 family is a good upgrade if you want specific new features, such as uNFS, Swift API, and HDFS enhancements.

 

If you are running a version of OneFS prior to OneFS 7.0, we encourage you to upgrade to a Target Code version of OneFS prior to June 30. For a list of recommended OneFS Target Code versions and release notes, refer to the Target Code table on the Uptime Information Hub landing page.

 

OneFS Upgrade Phases

 

There are five phases to the upgrade—planning for the upgrade, evaluating the health of your cluster, completing pre-upgrade tasks, performing the upgrade, and completing post-upgrade tasks. To help you plan your OneFS 6.5 upgrade maintenance window, this article highlights important information about each upgrade phase. The OneFS Upgrades Info Hub features documentation links and product announcements that will ease and enhance your upgrade planning. To protect your data and help to ensure a successful upgrade, read the Isilon OneFS Upgrade Planning and Process Guide. If you need assistance at any point in the upgrade, contact EMC Isilon Support Staff or your account team. For quick answers to your upgrade questions, see the OneFS 6.5 Upgrade FAQ on the OneFS 6.5 Upgrade Info Hub.

 

Phase 1: Planning for the upgrade

 

Visit the OneFS Upgrades Info Hub to find links to all the required documentation that you should review when planning for the upgrade.

 

We recommend that you to engage the EMC Remote Proactive service to assist with the upgrade. The Remote Proactive team can manage your upgrade from start to finish. If you have a Premium Support contract, upgrade management by the Remote Proactive service is included in your service contract. Be sure to reach out to them at least two months before your scheduled maintenance window.

 

If you would like help with your upgrade but do not have a Premium Support contract, the Remote Proactive service is not included in your service contract. However, you can open a service request to obtain it. This option is available only through June 30, 2015. You can create a service request at https://emc.support.com.

 

If you choose not to use the Remote Proactive service, you must follow all of the procedures detailed in the Isilon OneFS Upgrade Planning and Process Guide. Visit the OneFS 6.5 Upgrade Info Hub for all the required documentation that you should review when planning for an upgrade from versions prior to OneFS 7.0.

 

Phase 2: Evaluating the health of your cluster

 

Evaluate the health of your cluster to verify that it is in a supportable and operational condition. This will help to prevent upgrade issues. Make sure your cluster is healthy by reading the EMC Isilon Cluster Preventative Maintenance Checklist. Resolve any issues before upgrading. If you have questions about the health of your cluster, contact your account team for assistance.

 

The upgrade from OneFS 6.5.x to a supported version of OneFS is a simultaneous upgrade process. We recommend that you make a full backup of data before you start, and, if needed, that you mirror your business critical data to a second site. Prepare your users for the cluster unavailability. Client applications may require restarting after the OneFS upgrade, and you might have to refresh mounts and shares.

 

The OneFS 6.5 Upgrade Info Hub lists specific conditions to check for, as well as relevant resources.

 

Phase 3: Pre-upgrade tasks

 

Begin your pre-upgrade tasks at least one month before the upgrade. Tasks include scheduling a maintenance window and reading the documentation on the OneFS 6.5 Upgrade Info Hub.

 

Be sure to read the release notes for the version of OneFS that you are upgrading to, as well as any maintenance release notes for the OneFS family that is best for your organization.

 

If you run custom scripts in your environment, read the OneFS CLI Mappings Guide to determine what command line syntax has been modified and to make appropriate preparations ahead of time.

 

Important hardware and drive firmware update considerations

 

If your cluster is running Drive Firmware Package 1.7 or earlier, you must update the drive firmware to 1.8 or later before upgrading OneFS. For details, see ETA 193819. For the latest firmware versions and release notes, see the Uptime Information Hub.

 

Drive firmware updates for OneFS 6.5, 7.0, and 7.1.0 require you to reboot each node individually. If you are not running the latest drive firmware package, and if your workflow can tolerate additional downtime, reboot your cluster prior to the firmware update. Rebooting the cluster might expose latent hardware issues, and unresolved hardware issues can cause the upgrade process to fail.

 

Note that if you want to upgrade to OneFS 7.1.1, and your cluster contains i-Series nodes, 7.1.1 will not be compatible with your environment. If you'd like to upgrade to OneFS 7.1.1, contact your account team prior to June 30, 2015.

 

Note that upgrades from one major OneFS family to another major OneFS family require a complete cluster reboot. This process is known as a simultaneous upgrade. A simultaneous reboot upgrades OneFS and restarts all the nodes in the EMC Isilon cluster at the same time. While the nodes are restarting, the cluster is unavailable for client connections. To learn more, refer to chapter 4 of the Upgrade Planning and Process Guide, "Completing a simultaneous upgrade of OneFS."

 

Phase 4: Performing the upgrade

 

First, read and follow the steps in the section "Completing a simultaneous upgrade of OneFS" in the OneFS Upgrade Planning and Process Guide.

 

Perform the upgrade steps outlined in the OneFS 6.5 Upgrade Info Hub.

 

Note that if you need to upgrade to the functionality in OneFS 7.2, you cannot directly upgrade from OneFS 6.5 to OneFS 7.2. The path to upgrade from OneFS 6.5 to 7.2 is to upgrade to OneFS 7.1.1 and then to upgrade to OneFS 7.2. We recommend that you schedule two separate maintenance windows for each upgrade to give the Job Engine jobs time to complete.

 

Upgrades, including necessary reboots, can take anywhere from 15 minutes to several hours.

 

Phase 5: Post-upgrade tasks

 

In this phase, you will complete post-upgrade restoration and change management tasks to help ensure that your cluster performs as expected. Examples of post-upgrade tasks include:

 

  • Resolve any events or alerts by running the isi status command and reviewing the cluster event log (CELOG)
  • Install patches
  • Review access control level (ACL) settings and re-establishing user privileges
  • Restore client connections and custom settings
  • Configure base directories for access zones
  • Reconfigure SMB shares, home directory templates, and HDFS settings within access zones
  • Verify Kerberos migration
  • Update SMB auditing
  • Re-install Aspera
  • Update customer scripts

 

If you need assistance at any point in the upgrade, contact EMC Isilon Technical Support or your account team.