Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
When a job fails and there are a lot of parameters to populate, it would be very handy to have a "Resubmit with Edits" option that lands you in the Run Workspace page and populates all parameters & settings as configured for that job. This would let you make any tweaks/edits before submitting again.
The first step in upgrading FME Server is to back up config files from the Server, FMEServerConsole and Web Services folders manually. This is time-consuming particularly for the web services folders, not least as it's a lot of propertiesFile.properties files with exactly the same name.I'd suggest that if you have a default installation of FME Server - so everything goes in c:Program FilesFMEServer - that backup of these files should be automated either to be triggered from FME Server itself (possibly included with the Backup process in step 2 of the upgrade, which backs up repositories/users etc.) or as part of the uninstall of the old version (so a checkbox along the lines of "are you upgrading and do you want to back up config files?" then a corresponding one on the new installer).Would make things quicker/easier and cheaper.
***Note from Migration:*** Original Title was: Improve overwrite existing items when importing projects - sharedresourcedata folder options FME server > manage projects > import a project has super useful functionality to overwrite "items". When exchanging projects between servers, there can be differences between subfolders. For example $(FME_SHAREDRESOURCE_DATA)FolderYearMonth project files may differ between servers. Having additional functionality to choose which items in these folders would greatly assistance projection migration data management
FME Desktop 2022.1: Having a functionality to be able to click the gear sign of a transformer without getting the displayed table disappeared, ElaborationFME Desktop 2022.1: Why clicking the gear sign of a transformer gets the displayed table disappeared?
In Automatons it looks like Event as Json has the following:"automation.id":"workspace":"timeRequested":"requestKeyword":"timeStarted":"source":"repository":"jobsuccess_topic""LogFileName":"jobfailure_topic":"StatusMessage":"logFullPath":""StatusNumber":"logHome":""timeFinished""id":"time":"logUrl""event.id""automation.name":It would be nice to have NumFeaturesOutput as well, I can mine it out of the log file, or the job_hisoptry table, but it would be nice to just have it in the JSON
***Note from Migration:*** Original Title was: Named ports of collapsed bookmarks should show up in Data Inspector/Visual Preview If ports of collapsed bookmarks are named it is mandatory that these names are displayed in Data Inspector/Visual Previews "Data Control window" and "Feature Information window". Currently names of underlying Transformers are displayed which causes confusion during inspection.Situation in Workspace:Situation in Data Inspectors "Data Control Window"Situation in Data Inspectors "Feature Information Window"
Currently, when attempting to update workspace published parameter values for an Automation, if the parameters contain any password fields then no parameters can be updated without also re-entering the password values.It would be good if other parameters could be updated without requiring all passwords to be re-entered.If this is not possible then at least the error message could be updated to be something more friendly than the current: "javax.crypto.IllegalBlockSizeException: Data must not be longer than 245 bytes"See these two related postings:https://knowledge.safe.com/questions/93897/javax-error-when-saving-fme-server-automation.html?&childToView=112885#answer-112885 https://knowledge.safe.com/questions/104498/fme-server-automation-error-javaxcryptoillegalbloc.html?smartspace=automations-fme-server
There are those of us who create workspaces as products/tools which are distributed to customers who themselves have FME.In the development process it's nice to be able to split up parts of a workspace in to modules (linked custom transformers) which can be worked on independently - A very standard approach for developing with code. Using this method makes working with FME together with git a little more easier to handle - changes can be made in separate branches and merged without conflicts.There is a big downside to Linked Transformers which is that the end user experience is not as nice and requires additional work either by the customer or via another method to make sure all definitions of Linked Customer transformers are up-to-date. Better is to have the Linked custom transformers embedded prior to delivery.Currently there is no method to embed custom transformers except to open the workspace and choose embed for every one. This is not an option for automated pipelines. A method to embed all Custom Transformers would really help.
Often workspaces can't be run because the FME author created a new repository but forgot to "Share" as the last of the publishing process (step1-6 is FME form/desktop and step 7 is only possible in FME server/flow).So I'm asking for a popup/reminder to FME admin that a repository has 0 user(s) and 0 roles(s).What would be even better would be to set a default permission set. For example when adding a new repository, ask the user if they want to set recommended defaults of fmeadmin role has "Full Access" by defaultfmeauthor role = "Can Download" and fmeuser "Can Run"
Add the ability for an automated process to be triggered at a given interval over a specified time period (i.e. every hour from 0800 - 1700 daily). This would allow processes requiring quick response times to run while people are working and resource intensive processes to run at off peak hours.
Currently, it is only possible to download a single workspace version from FME Server web interface through the version history. For backup purposes it would be helpful if you could choose to download all workspace versions (commits/non-commits) for all repositories, for single repositories and for specific workspaces.See also this post: https://knowledge.safe.com/questions/111456/backup-all-fme-server-workspace-versions.html
Workbenches that produce a lot of disk io will perform poorly on Azure Kubernetes Clusters that are using Standard HDD OS disks.Because of this; It would be really nice if you can configure the FME_TMP environment in the values.yaml file.As solution for these clusters; you could then attach a new data disk with a better performance on your virtual machine scaleset, and set the FME_TMP environment to this disk.This would be also really helpfull if you want more flexibility in the size of disk. Otherwise
It would be nice if we could easily rollback an FME Server Project or backup restore action. As a partner who offers both support and training I assume we do a bit more of those kinds of actions than the average FME Server user but it's a bit of a hassle trying to go through all the items to see which ones can be safely removed (obviously we use a test environment for that, but still)
As English sometimes disturbs the users, it would be great to add the ability to either change the language or disable the help when drawing with geometrypicker.
Descriptions are useful for FME Server Apps, but I would like to be able to provide an internal/private readme for my Apps that is not visible to the end user. This way I can track dependencies and other metadata that will help me manage my app, but that I don't want to expose to end users.
It would be a great feature if users wouldn't need to click the 'Use Windows Credentials' button when they call a Server Apps which 'Require Authentication'.Users are often not interested in the underlying system, they just want to carry out tasks conveniently.The automatic login would provide a sleek user experience, but still in control who can do specific tasks.
It would be nice if the selected app (form) in the gallery would open in a frame below the app icons instead of in a new tab when selecting an app. The user could then navigate between the app in a much nicer way.
One of the great things about the FMEServerJobSubmitter is that it lets you define the connection (and hence the user) that the workspace will run under. The new Automations do not allow for this.It would be great if a Workspace action in an Automation could provide a "run as" option (perhaps under the Advanced parameter settings).
It would be helpful if FME Server for Linux could support running the component services as a service account tied to Active Directory(AD) or similar identity store rather than just as a local account created by default. It would be ideal If this could be set during the installation process as it is during Windows Distributed/Fault Tolerant installations. Our Enterprise utilizes both MS and Linux environments extensively and I have encountered issues integrating across platforms due to User security issues. Services such as PBIS make AD integration fairly seamless on Linux distributions so in theory this should be achievable.
***Note from Migration:*** Original Title was: Ability to download data files stored in the 'workspace repository' form the GUI of FME Server In the GUI of FME Server, when you go to 'Workspaces', and then select a repository, you get to see a table with all workspaces in that repository including some information about those workspaces. In particular there is also a column 'Files' that indicates if there are data files stored for that workspace inside of the repository (i.e. the files that can be uploaded to the repository of the workspace when publishing a workspace from FME Desktop to FME Server).If you click on 'Files' in the FME Server GUI (in the row of the appropriate workspace), you get to the Workspaces > REPOSITORY_NAME > Files menu, and can see all files that are stored for that particular workspace (in the workspace repository). However, it seems that there is no option to download these files from this menu in the GUI of FME Server (nor another menu in FME Server), as there is only the option to select the Files to remove them. I thought this is really weird and besides a Remove button, I would have definitely expected to also have a Download option (button) here...There is still the option to download the files stored in the workspace repository when you download the FME Server workspace through FME Desktop (when you import there is the button to include data files if present). However, as mentioned above I couldn't see a way to do it in the GUI of FME Server, and I would think it would be logical/preferred if this would also be possible from the 'Workspaces > REPOSITORY_NAME > Files' menu in the GUI of FME Server. Curious to hear other people's thoughts on this. Would be great if this can be added in a future version of FME Server.Kind regards,Thijsps. we are currently using FME Server 2021.0.2 Build 21321 - win64
Hi all,Many existing FME Server 2018 (or previous) installations are, or will be upgraded in the near future, to 2019+.I think it would be very useful if existing Notifications could be converted / imported into Automations.
In line with this idea, FME 2019 supplied the ability to auto-color annotations based on the bookmark it is placed in. Very helpful.However, if I manually adjust the color of one of these annotations this auto-color functionality is lost. I would like to have the option to enable this again so that the annotation coloring is 'restored' to be related to the bookmark.
The side effect of having FME database on a separate server is that it must be rebuilt at every upgrade. The provided drop/createUser/DB/Schema is not bad but the process of the FME database rebuild could make life easier with a couple of improvements (which are possible in all SQL Server/Postgresql/Oracle):In addition to dropDB/dropUser an emptyDB which drops all the database objects in the createDB/Schema script but not the database itself would save the user management and would require less privileges.Using variables in the createDB script eg. create the db or not, starting value of job_id, etc
FME Server projects are already a great way to migrate workspaces and related resources between FME Server instances in a DevOps-style way, without having to manually re-publish through FME Desktop. This would be even more powerful with the introduction of deployment slots similar to how they work in Azure AppService, as it would allow updates to existing processes to be deployed and tested in the target site before replacing the live deployment. This would be particularly useful in environments where there is no pre-PROD instance available.It could work at the project level and essentially create a parallel namespace that resources get deployed to, and where they can have the same name as the live ones, but with different settings. As an example, a web service or database connection called 'MyDbConn' would point to a pre-PROD database while the same database connection in the live deploymen slot would point to the PROD instance of the same DB.Once everything has been tested and is working OK in the parallel namespace, the live slot and the parallel slot get (hot-)swapped to go live with all updates at once, ideally with no service interruption.If any issues were found afterwards, the old configuration could be restored simply by swapping back, and then updating/testing the resources in the parallel namespace until they work OK.
In-depth courses for all FME skill levels.
Explore our extensive collection of how-to guides and tutorials.
Log an issue with us. Our support team is ready to help.
