Problem with some Filetypes and celery/flower

[rusty9283]

Hello!

If I "index dir" sometimes the cpu goes to 100% and extracting is stuck so after some testing i have seen that the "problem-files" wich are extracting are from type ".STEP" / ".SLDASM" / ".SLDDRW" but not always.

So I have to go to flower ui and terminate all running tasks (8 tasks) and restart service "opensemanticetl" than the extracting goes on.

Is there a explanation why this problem happens?

By the way is there a possibility to have the filetypes like "STEP" in the index but prevent to extract them? Because we do not need an extraction of them.

Many thanks!

[rusty9283]

After some more tests I think it's definitely a problem with some "3D-Files".

• Index a directory with 1600 files and no 3D Files like STEP usw. --> Perfekt
• Index a directory with 600 files and many STEP files --> nightmare - I have to terminate most of the taks from flower.

I don't know why but I think to exclude theeses files fomr extracting would help but don't know how to manage.

Thanks!

[mosea3]

I also just had the Indexing Service stuck. Turned out I acidentally broke my config file.

First make sure your config file is allright. You may test this by indexing a specific file

opensemanticsearch-index-file /path/file.pdf

If you get a python error, that may indicate your problem.

The other independent machinery are the opensemanticsearch services. You have

1. Opensemanticsearch ETL
2. Opensemanticsearch ETL-Filemonitoring

(On Debian) you may check its health by executing: systemctl --full --type service --all

Start/stop/get status by systemctl service status opensemanticetl systemctl service stop opensemanticetl-filemonitoring

use service status to check whats going on around those 2 services.