Question

FME Server vs Desktop reading from SQL Server Spatial to Esri FileGeodatabase (API)

  • 26 December 2017
  • 7 replies
  • 5 views

Badge

I have a total of 4 million records from various featureclasses stored in SQL spatial database and I have one workspace reading from SQL and writing out to an Esri FileGeodatabase (using API, not ArcObject writer). When I run this workspace in Desktop, it took approximately 16 hours to finish. However, it took 3 days to finish in FME Server. I checked the timestamps of the log in FME server, writing out was gradually getting slower and slower. I was just wondering why there is a significant difference between Desktop and Server with the same workspace?


7 replies

Badge +2
Hi @kuanglee

 

I have a few questions about this situation. What version of fme workbench and server are you running? Where is the output gdb located in both cases? What are the specifications of each machine running fme workbench and server? Was fme server running other jobs at the same time on the same machine?

 

Badge
Hi @jlutherthomas
  1. 2017.1 for both FME Server and Desktop.
  2. server output to ...fmeserver/resources#/FME_SHAREDRESOURCE_TEMP and desktop output to my local folder.
  3. There was only one job running in FME server. Thanks.
Badge +2
Hi @jlutherthomas
  1. 2017.1 for both FME Server and Desktop.
  2. server output to ...fmeserver/resources#/FME_SHAREDRESOURCE_TEMP and desktop output to my local folder.
  3. There was only one job running in FME server. Thanks.
Do you know the hardware specifications of your two machines?

 

If you run the job again in FME Server does it take the same amount of time?

 

 

Badge

Hi @jlutherthomas

My desktop:

Windows 10 Enterprise; IntelXeon 3.5GHZ; 32 GB RAM, 64-bit OS.

FME Server: FME Server on Docker inside Ubuntu Server VMs.

Server Hardware:

Windows Server 2016

CPU: 2x Xenon e5-2620

RAM: 198gb RAM

Disk: 10k SAS drives not sure of the raid config

Badge

Hi @jlutherthomas,

Re-running caused the same amount of time.

Badge +2

Hi @jlutherthomas,

Re-running caused the same amount of time.

Ok, I'm really not sure what's going on.

 

Are you able to report a problem here.

 

 

If you could give us the workspace, data and a link to this thread that'd be really useful and an expert will be able to investigate.
Badge +2
Ok, I'm really not sure what's going on.

 

Are you able to report a problem here.

 

 

If you could give us the workspace, data and a link to this thread that'd be really useful and an expert will be able to investigate.
Also @kuanglee log files with timestamps would be great, for both the desktop and server runs

Reply