FME Flow

Oh, no! The lights have gone out! The data center is on fire! Execute the Disaster Recovery procedure right now!! ...But how do we handle all things FME?  This is an opportunity to share your FME experiences while setting up your Disaster Recovery environment! What are some of your best practices and gotchas that you can share with the FME Community?  How do you handle all the IP address changes? What if your clients connect to servers via hostnames?  (FME Desktop) How do you handle switching (floating) license servers? What if clients are unable to access the FME Desktop Licensing Assistant?  (FME Server) How do you ensure that all your workspaces and source data are available in the new environment?

Afternoon All, i have read the documentation on 64bit FME vs 32 bit FME but am still a bit lost... I have a complex 32bit fme workspace in 32bit 2015 FME on windows 10 64bit. Its reading ESRI API file geodatabase data and writing ESRI API file geodatabase data and failing because of insufficient memory issues....My 64bit PC (windows 10) has 64gigs of RAMM and obviously the 32bit FME is only addressing 3 gigs of this. Am i likely to find my workbech will work (get past the insufficient memory issues) if i run it on FME 2015 64 bit rather than FME 2015 32 bit.  Thanks Steve

I have a publication that monitors a folder for new files that are created.I also have a working workspace that takes any number of files from a network location and uploads them where I they need to go. When trying to set up a subscription I get the following errorI have been able to run the workspace on its own however it fails when initiated by the topic.I can see the SourceDataset_PATH parameter is missing from the log but.Published Parameters Parameter Value SourceDataset_PATH FME_TOPIC_MESSAGE$(FME_SHAREDRESOURCE_SYSTEM)temp/subscriptions/277a9c3190634de39102c05756666082.json FME_TOPICDatasetWatchstatusMessageParameter 'SourceDataset_PATH' must be given a value. Any ideas?

Hi all, I'm trying to set up a workspace that should fire every time a file is added to a folder. I came across the reader format "Directory Watch Reader", but so for I have been unable to find any documentation about how to use this reader type. If this is covered in the documentation anywhere, or in an article here at knowledge.safe.com, I'd be grateful for a link.

As the title says, I'd like FME to look at the floating license of another software, read into who is using a license and then send those users an email asking them weather they need to be using a license/politely telling them to give me their license!Thanks,

Why does Topic Monitoring work without error in Google Chrome, but when I try to enable it in Mozilla Firefox, it starts and immediately stops?I have configured my FME Server installation to use a self-signed certificate, but I have already made the security exception for each web browser! What gives?Note that the fmewebsocket.log doesn't contain any helpful information, it only lists the attempted starts and immediate stops:Fri-21-Oct-2016 05:20:17 PM   INFORM   New I/O server boss #5   408506 : channelOpen => -1,008,910,118Fri-21-Oct-2016 05:20:17 PM   INFORM   New I/O worker #1   408511 : Channel closed: -1,008,910,118Fri-21-Oct-2016 05:20:23 PM   INFORM   New I/O server boss #5   408506 : channelOpen => -231,922,873Fri-21-Oct-2016 05:20:23 PM   INFORM   New I/O worker #2   408511 : Channel closed: -231,922,873Fri-21-Oct-2016 05:20:23 PM   INFORM   New I/O server boss #5   408506 : channelOpen => 2,132,443,954Fri-21-Oct-2016 05:20:23 PM   INFORM   New I/O worker #3   408511

Hello,The window named "Transformer Gallery" is very big. How to how to reduce the size and move it at the bottom of the screen

Hi all,I do have a question about how FME and S3 interact each other. Im working with big amount of data that will likely increase very fast (terabytes). In my way to test how could we efficiently store the data and later work it with FME Im doing some tests basic tests with S3 from amazon. My point is that I can store those terabytes in Amazon but when I try to work with them with S3objectlister and S3downloader I found that you need to really download data to process it so all in all you will need to have extra storage in your server or desktop. So for example if I want to work with 5 gb of cloud data that I have stored in S3 I need first download this on local and then work normally on FME? I have not found a way to connect S3downloader directly to feature reader so I dont need to download all the data and process on the fly: Not sure if I explained correctly or not, or if Im missing somehting in the FME documentation that could do in other way...

I have .dwg file with large number of lines which are snapped to points, so they form closed polygons. I would like to create closed polyline from snapped lines so i can calculate area of polygon, and the main criteria is that lines are snapped to node of point, which is not always the case (lines are often snapped to endpoint, nearest point, or not even closed). My question is: how can i detect where lines are not snapped to nodes of points, so i can manually snap them? Thank you for answers.

I see over a dozen FMEPublisher.exe and the same for FMESubscriber.exe? Even when no jobs are processing.

I have been using this example: https://knowledge.safe.com/articles/1117/data-distribution-with-web-maps.htmland building my own verison using our own data. I have been asked if it can stream the data to the web map once the button has been ticked for a certain dataset so the user can see the data before they request it. I.E. in the example above, once you have drawn your polygon, you click on BusRoutes and the bus routes inside of the polygon would then draw up on screen so the user can see what they are about to download.Can this be done?@aaronkoning @LauraAtSafe

Hi,  I've signed up the FME Cloud Beta, and started to learn how to use it. As my first trial, I uploaded a workspace which transforms XML dataset into raster and writes the result into Geotiff file. Service type is Data Download. The workspace worked fine, and I was able to download zipped result file from the web page which was displayed after the completion, if I didn't check "Email Result" (Advanced Option) on the Configure page.  However, when I checked "Email Results" and specified my mail address, I was not able to receive the e-mail even though the job has been completed successfully.   I checked if the mail address was correct, and tried several times. But no mail has come ... Do I need some server settings to enable the "Email Results" option? Thanks in advance.  Takashi

Hi,I was wondering, is there any option to receive snmp message (trap message) on host: localhost, port: 162. I've tried connecting to TCPIPReceiver - gets me fatal error; tried get method in HTTP Caller - gets me wrong message (since no port is defined). Was trying to also define UDP publication in FME Server (Topic monitoring shows me nothing).Of course, I might be missing some crucial parameter. Might be, even, entrying wrong values to necesary parameters. So, please, correct me.Thank you.

Hi there,  We are working with FME Server for the past four months, and we've noticed that sometimes, seemingly out of the blue we get an error like this one:  Mon-24-Dec-2012 06:44:08 PM   ERROR    Engine-Thread-<ServerName>-Engine1   401854 : No message length returned in FME Engine reponse Mon-24-Dec-2012 06:44:08 PM   FATAL    Engine-Thread-<ServerName>-Engine1   401872 : Failed to get translation result. Returning failed result. Mon-24-Dec-2012 06:44:09 PM   WARN     Engine-Thread-<ServerName>-Engine1   402615 : Job 8469 failed and has been resubmitted. This is resubmittal 1 out of a maximum of 3. Mon-24-Dec-2012 06:44:09 PM   INFORM   Engine-Thread-<ServerName>-Engine1   401936 : Disconnected FME Engine: Engine1 on host <ServerName>  We are using FME(R) 2012 SP4 20120926 - Build 12266 with only one Engine running on a 64-bit Linux environment with 4 Gb of RAM, and we are currently in a testing phase - meaning we send an average of more than 500 r

Hi,As of late I've noticed that we've got a lot of sprawl with the names of our repositories. Does anybody have any best practices/recommendations on how their organization creates repositories/topics/notifications and assigns permissions?One thought I had was to create a repository per application and also per department/organizational group as needed. But I'm wondering if anybody else has any other ideas/thoughts on the matter. In addition, my thoughts really only apply to repositories and don't take into account permissions/topics/notifications.If it matters, my FME Server install only runs on my intranet. I don't need to expose it to the public.Thanks!-Jeff

I've successfully have emails sent MS Exchange forward to FME Server. No problems forwarding the email to another address as backup, problem is some email have 1-10 attached files. FMESeverEmailGenerator Content: allows only one attachment from a list email_publisher_content{0}How I just cc an email with multiple attachments?

Web User Interface is timing out, not displaying data, giving a connectivity error. How can I narrow down the problem?

Do you use a proxy with FME Server? If so how do you implement it?

I've read through the documentation on email and can't find anything about this. When I pass a layer (ArcGIS) through the FMEServerEmailGenerator and the FMEServerNotifier I get an email for every record in the layer. If there's 68 records I get 68 emails. I understand how to get one email sent to me with statistics from the layer. That's not the issue. But what I want to do is basically have FME send one email that has the table of all 68 records (Along with the column headers if possible ... Here's an example: ) in the body of the email. Almost as if you opened an excel file and selected a range of cells, copied them, and pasted them into an email ... If that makes sense?

We are designing a conversion-job spanning multiple workspaces. We can run one after the other using notifications by subscribing the next workspace to the succes topic of the previous.We could design this much more beautiful if we could e.g. start workspace 'E' and 'F' when 'B' and 'C' are ready. But when 'E' has a subscription to the succes-topics of 'B' and 'C' this is evaluated as an 'OR'. 'E' runs when 'B' or 'C' is ready.It would be nice to evaluate the notifications a workspace is subscribed to with an 'AND' clause. You should do 'this', when you have 'this' and 'that'. Is there a possibility to do this using FME only? Ofcourse we could make our own 'notification service' using some database and python or C#, but as far as I can see, the only thing I am missing in FME-server is an 'AND' clause in the notification subscription.With such possibili

We have our servers running on RHEL 6, is FME Desktop available for this?Thanks,Vinod

I am using a C# REST service call to start a workspace. Passing the notification directives in the body.The notification subscription is setup with the PUSH service call. Basically, the subscription makes a service call when the workspace is failed or successful. When I published the workspace, with JobSubmitter, I made sure the right writer output is selected to be included in the service call, but the writer output will be included only if I run that workspace from WebUI, not with the service call. Is there a directive\\parameter that I can use in the initial service call to tell the notification to include the writer content in the notification body?Any thoughts?

Dear FME community,  is there a way to connect to an linux/ubuntu server from within FME Desktop? We have a server which has a specific IP adress. Say e.g.: 123.45.67.89  On this server which is a virtual machine runs Ubuntu. We would like to read from this server and if possible update the files in a specific linux folder. Is there a way from within FME to gain access to an ubuntu server?  Kind regards Thomas 

HiWe have a problem running scheduled tasks when we activate AD-authentication on our FME Server 2015. We get the following message in the log:MEServerJobSubmitter(ServerJobSubmissionFactory): Could not initialize connection to server "localhost". FME Server error: `Authentication failed: Failed to bind to Active Directory with the given username and password.'Anyone got this to work ?Kind regardsGuha,Stockholm Water

Hi,I'm looking at ways of assessing capacity of our fme server instance(s), and as well as regular things like cpu/ram capacity on the server I'm interested in the application specific metrics.In particular does anyone have a workbench/tool that will query an FME server 2015 repository and report the max number of engines running in a defined time period, lets say hourly. (I'm interested in the concurrency of engines, so I can reduce the chance of queuing, particuarly when I have active time dependant jobs.)If any one has any other stat gathering workbenches, I'd be interested in them too. I have some that query the repository and work out some stats per workbench, avg/max run times, success/fail counts that sort of stuff.Thanks in advance,Nick