With many Connect:Direct users transferring tens of thousands of files every day on their IBM Managed File Transfer (MFT) platform, the solution is now considered mission critical infrastructure. You count on it to work behind the scenes to distribute information internally across your enterprise or externally to and from customers. Connect:Direct powers your business, so you need to make sure it is always up and running. The challenge though, is that hardware fails, operating systems have issues and maintenance must occur.
Operating Connect:Direct in a high availability environment ensures your file transfer capabilities are always on and available 24x7x365, even when there is an outage, planned or unexpected. By building redundancy into your Connect:Direct service, you can continue to transfer files while you resolve problems, take care of maintenance and bring systems back up. Highly available architectures allow these activities to occur with no impact, or even awareness, to your trading partners.
How it works
Historically, high availability (HA) has presented some challenges in terms of additional hardware complexity and maintenance. Your IBM MFT team has simplified how to build and manage HA environments for either of the two options you decide to pursue: the traditional active/passive model or an active/active approach.
Active/passive. In this model, one node handles all production needs and a second remains inactive until the production node fails or needs to come offline for maintenance.
Active/active. Here, the nodes sit behind a load balancer that distributes file transfer activity across a cluster containing multiple instances of Connect:Direct [see Figure 1].
Figure 1. Connect:Direct deployment in an active/active environment
The cluster of instances appears as a single node to the outside world. All nodes are active, so no hardware is sitting idle. The load balancer periodically runs system health checks and in the event of an issue, or while maintenance is performed, will automatically redirect traffic to nodes that are operational. Ideally, at least three nodes are behind the load balancer so that if one goes down the others can handle the traffic. This also provides scalability by allowing you to adjust the number of instances in your cluster.
NOTE: Active/active is not available for Windows environments.
An active/passive cluster requires more hardware than a single node, as the second node remains on standby, while an active/active approach requires a load balancer and software that active/passive doesn’t. However, many organizations already have a load balancer, and the automated failover, scalability and extra peace of mind often makes active/active a more attractive option.
What’s in it for you
- Meet SLAs. Downtime is not only disruptive, but costly, particularly when you have SLAs that include financial penalties if data isn’t transferred within a certain time period. Costs can run upwards of $100,000 per hour or more in some cases, including penalties and potential loss of business.
- Simplify maintenance. HA environments provide flexibility for scheduled maintenance, giving you greater opportunity to optimize system performance which can also extend the life of your infrastructure investments.
- Minimize fire drills. Seamless failover allows you and your staff to stay focused on business as usual, despite an outage.
How to get started
To start building your high availability Connect:Direct environment, contact your IBM sales rep or business partner. They’ll be happy to arrange a review of your environment and work with you to map out a high availability/disaster recovery plan.
I encourage you to post any questions in the discussion forum and we’ll answer them quickly.
Finally, if you haven’t explored Connect:Direct V6.0, our major new release that became available on December 7, 2018 you can read more here. Watch a demo of the new user interface featured in this 30-minute webcast, “Diving Deep into the New IBM Connect:Direct V6.0”.