Each year at this time we send out a survey asking for feedback on FME Server. Your feedback is crucial in shaping our focus for the coming year. In combination with existing enhancement requests that you’ve provided, the survey results directly impact efforts in development, knowledge base, marketing, education and documentation.
This year I’m asking you to again contribute your thoughts and ideas:
My promise to you is that I will work steadfast to ensure that your needs are addressed. You will be directly responsible for making FME Server better!
What the heck happened to my feedback from last year?
With over 17% of FME Server customers responding, the results were:
In the remainder of the post I’ll let you know how we addressed your feedback in the latest release and how we plan to continue to address it in subsequent releases.
Most Improved for You
Let’s start with the Web User Interface. For 2015 it is brand new. It looks better, it’s more intuitive and it is responsive, meaning it will work on desktop, tablet and mobile browsers. We’ve also added tooltips and other small tweaks to enhance the user experience.
The Notification Service is a powerful, flexible tool for automating workflows, but you said it shouldn’t take a PhD in rocket science to use it, and I agree. For 2015, we’ve made a number of small changes that really add up.
For example, if you are creating a publisher or subscriber you can now define the topic at the same time. The topic monitoring functionality, which makes it easy to verify that publishers are working, has now been moved to the notification page. It’s also quite easy to verify that subscribers are working by using the test topic functionality.
In a recent blog post I spoke about how Webhooks are great for simple server-to-server communication. If you think about it, topics really are FME Server Webhooks. Each topic has a corresponding REST API call; a Webhook callback URL. Figuring out the URL is now trivial since it is displayed when you use the test topic button.
File Geodatabase is a tricky format to say the least. It’s actually a folder and there are very few formats like it. Folder or not, you asked us to be better and FME Server 2015 is up to the challenge. You can now browse and select File Geodatabases in the web user interface, publish them alongside a workspace. And due to some advanced functionality offered by Google Chrome, you can even upload an entire File Geodatabase folder through the web UI.
Most Important to You
Reliability and Availability are really intertwined concepts. You want a system that is reliable enough not to break, but if it does break somehow, there should be a mechanism to maintain service availability. We’ve always worked on both of these topics, but given your feedback we really ramped up our efforts for 2015.
Reliability. One of the most annoying things about any enterprise system is when the disk fills up. Quite often it is due to log files or temporary data and we have dealt with both of these two cases. First, we made it so that all unneeded temporary data is deleted at a regular occurrence. Second, we’ve introduced a new bit of functionality that deletes log files after a certain age. The age of the files to delete is configurable and you can also monitor and delete files from user defined directories; configuration is done in the web UI.
We’ve also enhanced the FME Engine to be more stubborn… I mean resilient. If it can’t get a license or connect to the FME Server core it won’t give up. It will simply pause for a minute and then try again. This can help with network interruptions or cases where you legitimately need to momentarily stop the License Server or FME Server core.
We’ve done a few other things to make it more reliable as well, but should there be an unforeseen problem… well we’ve got that covered. FME Server has always provided a high availability configuration, but this year we present two configurations optimized for your requirements (basic and advanced).
We’ve spent quite a bit of time clarifying both of the high availability configurations in the Administrator’s guide. Basic high availability is a simplified version of what we documented last year. Advanced high availability covers the case where you want to distribute job processing (i.e. FME Engines) across multiple machines. For that configuration, we made it so that the FME Engines will failover, switching from the failed FME Server core to the secondary core. If you need to install FME Engines on separate machines for any reason, you’ll be glad to know that failover configuration is a good option.
FME Server adds a very powerful and flexible tool to your organization’s toolbox, but admittedly it can be a challenge to work with given the variety of uses. The survey results indicate that the Knowledge Base was far and away the number one resource for learning. Given that, we spent the last year on a bit of a Knowledge Base writing spree producing 10 getting started pages, 14 tutorials, 10 web examples, a troubleshooting guide and a known issues page. We also did some small things like adding links to documentation from previous years.
Each journey begins with a single step
Now what about the rest of the stuff that’s important to you?
- Sharing and Delivery of data
- Cloud data accessibility
We didn’t disregard it. In fact, we have laid the foundation for these things in the current release.
FME Server is automation. You may have heard that message on the FME World Tour, and we mean it. We have already started overhauling the experience of creating workflows with FME Server. Workflows which contain: a trigger, a workspace and a response. Workflows we term Automations. The complexity and confusion of publishers, subscribers and topics is to be hidden. We’ve already added the API for this, and the new web UI is engineered towards this transition. Scheduling is an automation as well. So that experience will benefit from standardized access to job options and response configuration.
In terms of sharing data and cloud data accessibility, we are definitely seeing the need to read, write and monitor cloud data services. We’ve laid the foundation here as well. FME workspaces now have the ability to read/write OAuth web services and it is just a matter of time until FME Server can support these workspaces. The ability for the web UI to browse and select from cloud data services will round out the experience.
Circling back to automation, if your data is being added to cloud data services, then at some point you will want to automate workflows that use it. For this release we added the ability to monitor an S3 bucket for new files and run a workspace when detected. It also turns out that a workspace can be triggered when new files are available in services like BOX.com, Dropbox and Google Drive using FME Server’s Webhook support.
If you read this far, then you must really be interested in FME Server! So why not:
Ten minutes of your time can help shape product focus for a whole year. Thanks!