r15 - 2014-09-23 - 03:07:00 - Main.lfrankelYou are here: TWiki >  Deployment Web > DeploymentInstallingUpgradingAndMigrating > ServerRename > ServerRenamePlanningFor

Planning for server rename todo.png

Authors: LisaFrankel
Build basis: CLM products, version 4.0.1 and later

Page contents

Before you perform a server rename, review the overview information on this page, learn about supported and unsupported scenarios, and perform the prerequisite rename steps.

Important: Server rename is a complex and potentially disruptive operation because correcting the stored links to the server from other applications and systems can be difficult or impossible. Server rename is supported only for a specific set of scenarios and requires careful planning. Use server rename only as a last resort when other approaches are unworkable. To enable server rename, you must obtain a feature key file from IBM Software Support. When you contact IBM support, mention that you are requesting a "Server Rename feature key file". The key file is named ImportURLMappings.activate. Once received, copy the file to the JazzInstallDir/server/conf directory for the applications that you will rename.

Server rename

A server rename is defined as changing the public URL for the Jazz™ Team Server and one or more of its registered applications, after those applications are deployed. The URL change can include any or all of the following components: protocol, hostname, domain, port, or context root. An example of a public URL for a Jazz Team Server would be the string https://clm01.mycompany.com:9443/jts.

The Jazz Team Server and CLM applications use links to store relationships between stored artifacts that may span applications or systems, and to communicate with each other and with non-CLM applications. Server rename remaps existing URLs to new URLs to preserve the integrity of most links. Without the remapping, those links would be broken.

Note: There are many links that are outside the control of the Jazz Team Server and the CLM applications. The following types of links are not covered in the rename and would be broken:

  • URLs embedded in email, presentations, and documents
  • Bookmarks in browsers
  • Free text links that users paste or type into UI fields
  • Links from third-party applications

Risks and precautions

Performing a server rename is not without risk, particularly if the CLM deployment includes integrations to other non-CLM applications. Any change to a host name has the potential to break links in the data or to prevent CLM products from operating correctly.

If it is necessary to make changes to your topology, the preferred approach is to always do so in a way that maintains a stable URL, such as using DNS to route the host name to a different machine, or using a reverse proxy or virtual hosts. See Techniques for changing your topology for details. Where this is not possible, you can use server rename.

The following precautionary steps are recommended before you perform a server rename:

  • Plan your deployment carefully. See Deployment and installation planning for the Rational solution for CLM, particularly the topic Planning your URIs.
  • Be careful throughout the renaming process to enter URLs correctly, double-checking to avoid typos in host names, ports, or context roots. Some of these typos are not detectable by the rename process. To correct mistakes, you may need to perform another rename, or in extreme circumstances, you may need to restore from backups before performing another rename.
  • Prepare and thoroughly review the mapping file in advance of the actual rename.
  • It is imperative that you fully understand your deployment prior to a rename, specifically what external references and linkages you have to other production systems. Review the information in Supported scenarios for using server rename to understand the scenarios that are supported. Review the information in Impact of server rename on the Rational solution for Collaborative Lifecycle Management and Impact of server rename on integrated products to see if the integrations in your deployment are supported.
  • Carefully consider clients, such as the Rational Team Concert Eclipse or Visual Studio client, that rely on the CLM deployment to be renamed. To prevent any disruption and to preserve availability, notify end users of those clients in advance that you are going to change the URLs of the CLM servers used by those clients. Encourage end users to perform backups and to complete tasks in progress, such as code deliveries, before you start the rename procedure.
  • During the online portion of the rename, users may be able to log in to the server while the administrator is validating the renamed data. At this time, any operation that is not a read-only operation will result in an error. Ensure that users are aware of this behavior in case they log in before the online rename is complete.

Software version requirements

Before you can perform a server rename, you must upgrade to the required version of the CLM software. The following table lists the software version requirements for performing a server rename:

Table 1. Required software versions

###Add table here

For a complete list of system requirements for the current release, see System requirements.

What parts of the URL can be renamed?

Renaming a Jazz Team Server and/or the CLM applications requires a remapping of the URLs that are stored in CLM resources. You can rename the entire URL prefix, which includes all parts of the URL through the context root, but does not include the context path of a resource. Specifically, you can rename the scheme (or protocol), host, domain, port, and context root.

For example, suppose you want to rename the following old URL in a pilot deployment, which is moving to a centrally managed data center:

  • Old URL: http://bad.host.example.org:9443/ccm
    This URL uses an unsecure protocol (http), the default port configuration (9443), and a default context root (ccm).

  • New URL: https://good.host.example.org/ccm14
    This URL uses a secure protocol (https), changes the host name, uses the default port (unspecified in the URL), and a custom context root (ccm14).

Table 2. Parts of the URL that can be remapped in a server rename

###Add table here

Supported scenarios

You can rename the server to move a pilot deployment into production or to move an existing production deployment to new hardware. The server rename operation moves all existing projects and artifacts from one deployment to another. The operation does not support a selected project move function; that is, you cannot move only selected projects when you rename a server.

Read the description of the supported and unsupported scenarios below. See Server rename process example for the high-level steps in performing a server rename. See Software version requirements for information about the requirements for each scenario.

Setting up a test staging environment with production data

This scenario allows you to create a copy of an existing CLM deployment for test purposes only. In this scenario, when version 4 is running in production, you can also install version 4 in the test staging environment and copy the data and configuration files from production to the staging environment. You then use server rename to change the URLs in the staging environment.

This is useful for trying out new features or functions with real data without impacting the production database. One such example might be trying out a change to your process configuration or Rational® Team Concert work item type definitions. This scenario requires that the production server and the copied test server must never cross-link.

In addition, it is important not to connect a Rational Team Concert client to both the production and staging repositories. Always use a staging workspace for connecting to the staging server.

For additional details, see Topologies and mapping files for setting up a test staging environment and Setting up a test staging environment with production data.

Important: This scenario cannot be used to stage an upgrade from version 3 to version 4, which requires an isolated network where the URLs remain constant before and during the upgrade procedure. For details about this upgrade scenario, see Staging a test environment for the upgrade process.

Moving a pilot deployment to production

This scenario starts out as a small, pilot deployment that has been set up for evaluation purposes. The pilot has a limited number of users, who test-drive product features and create data. After the evaluation period, the goal is to scale up, add more users and data, and not lose any existing data that was created during the pilot.

For several reasons, such as a need to move the pilot to more robust hardware or a need to comply with corporate naming standards, a server rename is required. The pilot-to-production scenario must meet the following requirements:

  • The pilot must be a single CLM deployment (Jazz Team Server with the registered CLM applications) with no linkages to other CLM deployments.
  • All user clients and build engines must be at version 4 or higher prior to the rename.
  • After performing the server rename, the pilot deployment must be permanently taken out of service.
  • No integrations to Rational or third-party products are allowed, except for integrations to Rational ClearQuest®, Rational ClearCase®, and Rational Reporting for Development Intelligence (RRDI). RRDI can be moved from the pilot environment to a separate system in the production environment. Rational ClearQuest and Rational ClearCase must be production systems that can communicate with the CLM deployment in the pilot environment. At this time, moving Rational ClearQuest or Rational ClearCase from a pilot system to a production system is an unsupported scenario.
  • You must not be using the Derby database, or must migrate off of the Derby database prior to the rename.

For additional details, see Topologies and mapping files for the pilot-to-production scenario and Moving a pilot or full production deployment by using server rename.

Moving an existing production deployment

Starting from version 4.0.1, you can perform a server rename on a CLM deployment that is in full production. You can move your deployment to new hardware or perform a rename in place on the existing hardware. This production-to-production scenario must meet the following requirements:

Mapping file for server rename

Description of a sample mapping file

Dummy mappings to protect production data

Using a simplified mapping file

URLs with default ports

Case sensitivity

Additional URLs file

Errors during mapping file generation

Verifying a mappings file

Server rename on z/OS

Server rename process example

Topology diagrams and mapping file examples

This section provides example topologies and mapping files to help you perform a server rename. Use the example that closely matches your deployment environment.

Setting up a test staging environment

A staging environment is a test sandbox that includes a snapshot of production data isolated from the production environment. When planning a test staging environment, you need to be especially careful not to contaminate production data or vice-versa. You accomplish this by masking out production servers in the mapping file. Be sure to consider any additional CLM servers, or any integrated servers, that might be connected to the production server.

Starting Point 1: Single-server CLM production deployment

Copying the production environment

Starting Point 2: Production environment with two linked CLM deployments

Copying a single, distributed deployment to a staging environment
Copying both distributed deployments to a staging environment

Pilot-to-production scenario

Starting Point 1: Single-server CLM pilot deployment

Moving to a single-server CLM production deployment
Moving to a distributed production deployment

Starting Point 2: Single-server CLM pilot deployment with integrations

Moving to a single-server CLM production deployment with integrations

Production-to-production scenario

Starting Point 1: Single-server CLM deployment

Copying the single-server production deployment

Starting Point 2: Distributed production deployment with integrations

Renaming the distributed production deployment
Moving a single application in a distributed deployment to a new server

Starting Point 3: Two distributed deployments that are linked together

Renaming one of the linked deployments

Related topics: Deployment web home, Deployment web home

External links:

Additional contributors: None

Edit | Attach | Printable | Raw View | Backlinks: Web, All Webs | History: r15 < r14 < r13 < r12 < r11 | More topic actions
 
This site is powered by the TWiki collaboration platformCopyright © by IBM and non-IBM contributing authors. All material on this collaboration platform is the property of the contributing authors.
Contributions are governed by our Terms of Use. Please read the following disclaimer.
Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.