Deployments Timeout If Encryption Password Supplied for Non-Encrypted File

It appears as though deployments on unencrypted files fail with a rather long timeout of 30 minutes if an encryption password is specified. Specific error message is:

Build server timeout: Timeout exceeded

Edit: I should also add that the build .zip file in /outbox remains at zero KB this entire time; that was an interesting clue.

Armed with this critical knowledge, we will be silly and try to specify an encryption password on unencrypted files, but I wonder if OttoFMS might be improved to catch this and report the issue.

Running OttoFMS v4.9.0.

Hmm that is strange Matt. I’ll look into this and catch it if we can. Are you running installs/replaces or migrations for this deployment?

-Kyle

Hey Matt,

I’ve been unable to replicate this in my testing. What version of FMS and which Operating system is your source server?

Source server: Windows Server 2022 Datacenter
Destination server: macOS 13.7.1

Both servers are running FMS 20.3.2 and OttoFMS 4.9.0.
We were running the Migrate operation, but the first one we tried was a Replace (this is the one that timed out after 30 minutes).

Here’s the actual sequence of attempts (operation / error message / duration):

  1. Replace / Build server timeout: Timeout exceeded / 30:25
  2. Migrate / Build server error. http Status: 409. FMS Error 18000: build already in progress / 00:09
  3. Migrate / Deployment was aborted / 01:27
  4. Migrate / Build server error. http Status: 409. FMS Error 18000: build already in progress / 00:09
  5. Migrate / Deployment was aborted / 26:51
  6. Migrate / Build server error. http Status: 409. FMS Error 18000: build already in progress / 00:11
  7. Migrate / Build server error. http Status: 500. FMS Error 18000: error starting build ENOTEMPTY: directory not empty, rmdir ā€˜C:\Program Files\FileMaker\FileMaker Server\Data\Backups\OttoFMS\outbox\jit_build_8E8pkG4DBoO5’ / 00:10

The build .zip files that were generated with deployments 1, 3 and 5 were zero KB in file size. It seems that aborting these attempts did not free up OttoFMS from thinking that it was creating a build, 'cause the very next attempt would fail with ā€œbuild already in progressā€ā€¦ but once THAT failed, we were free to try again.

It was after #6 and before #7 that we identified the problem with the encryption password, and corrected that. Then we got the error message on #7 about the outbox not being empty, I restarted OttoFMS (which cleared out the outbox) and attempt #8 was successful and went beautifully.

Hey Matt,

All of those subsequent errors are a result of that initial build error. Restarting OttoFMS on your source server should fix it. I also have some fixes coming out in the next version that should make those subsequent errors not occur.

I’ll look into the encryption password issue and see if I can replicate it on one of our Windows servers. Thanks for all the info!

-Kyle

1 Like