Job continues after the “Deliveries: Folder” processor ended with an error

[DCCentR]

Job continues after the “Deliveries: Folder” processor ended with an error:

Снимок экрана 2024-11-27 160758.png (98.85 KiB) Viewed 2424 times

Or error codes like this not a sufficient error? Command executor processor after folder is set to execute on success

FFAStrans v1.4.0.7 & webinterface 1.4.0.89

Logs:

20241127-1312-5788-71e7-94fee41e1bab.zip

(8.61 KiB) Downloaded 140 times

Workflow:

Adv - копирование в импорт (nas04, nas05, nas06).json

(18.33 KiB) Downloaded 138 times

[admin]

Hi DCCentR,

I'm afraid this is one looks like a dreaded concurrency issue with the way we use file system as db in a farming environment. I would like you to make sure these registry parameters are set on each host and server hold the FFAStrans installation:

Code: Select all

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LanmanWorkstation\Parameters]
"DirectoryCacheLifetime"=dword:00000000
"FileNotFoundCacheLifetime"=dword:00000000
"FileInfoCacheLifetime"=dword:00000000

I investigated into these things years ago and they seem to be helping these kind of issues. In theory, they might also help with the stuck monitor files...and all db concurrency in general. If these values miss in you registry just add them. They are well documented online:
https://learn.microsoft.com/en-us/windo ... le-server/

-steinar

[emcodem]

that webint dark skin looks just awesome, i play with the thought about making it default or even dump support for light skin in near future

[admin]

@emcodem

i play with the thought about making it default or even dump support for light skin in near future

No you don't!

-steinar

[DCCentR]

Hi DCCentR,

I'm afraid this is one looks like a dreaded concurrency issue with the way we use file system as db in a farming environment. I would like you to make sure these registry parameters are set on each host and server hold the FFAStrans installation:

Code: Select all

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LanmanWorkstation\Parameters]
"DirectoryCacheLifetime"=dword:00000000
"FileNotFoundCacheLifetime"=dword:00000000
"FileInfoCacheLifetime"=dword:00000000

I investigated into these things years ago and they seem to be helping these kind of issues. In theory, they might also help with the stuck monitor files...and all db concurrency in general. If these values miss in you registry just add them. They are well documented online:
https://learn.microsoft.com/en-us/windo ... le-server/

-steinar

Thanks, I added those registry keys, I just need to find a moment to reboot the servers.
I'm not quite clear, do you mean this is meant to be the solution for my “Got error code 6” error or is it meant to solve the execution of subsequent processors when getting an error in the previous one?

that webint dark skin looks just awesome, i play with the thought about making it default or even dump support for light skin in near future

Yeah, it's cool Have you thought about turning on “Row Colors” by default? When I didn't see the highlighted red error lines I was a little confused

[emcodem]

Have you thought about turning on “Row Colors” by default?

We have row colors?
That one is an ever going topic, one year its default, the other it is not working at all, next year it is default off
I think it makes sense to turn row colors on by default, just as the dark mode should be on by default, but who am i to decide that

[admin]

I'm not quite clear, do you mean this is meant to be the solution for my “Got error code 6” error or is it meant to solve the execution of subsequent processors when getting an error in the previous one?

No, it's about making sure all i/o to the db folder is without any caching. This should will help concurrency issues. Other side effect of changing these registries can probably vary from system to system, and it might help other "strange" issues...but that's not the main reason. Turning of cache is normally never bad...as caching is after all cheating

-steinar

[emcodem]

Row colors are now default on again