[bug] ShareHound running for more than a week against one file server

[robertstrom]

Hello,

I am starting to play around with OpenGraph and ShareHound is the first tool that I am using to do so.

I pointed ShareHound to a couple of smaller file shares and it ran quickly and returned results (I have not fed those results into OpenGraph yet).

After performing what appear to be successful runs of ShareHound I pointed it to one locations file server and let it rip.

I started this ShareHound run on 11/24/2025. It is now 12/4/2025 and it is still running. I can run tcpdump filtered on the destination and I am still seeing SMB traffic and I am definitely seeing the numbers go up, so I do believe that it is still working. But ... this is a crazy long amount of time!

Two days after starting

Five days after starting

etc.

etc.

Have there been any other reports of collections taking this long? Is this what I should expect?

Has there been any testing on large file servers?

Suggestion, some sort of progress bar would be nice if possible.

[robertstrom]

It ended up not finishing. Ran out of memory. System has 12GB RAM

Another suggestion. It would be nice if there was a partial file that could be salvaged if something like this happens. As it is, as far as I can tell, there is zero output unless the collection is successful.

It might also be a good idea, if possible, to write intermediate results to disk so not everything needs to be stored in memory.

Any recommendations for the amount of memory needed?

[Sun Dec  7 18:11:09 2025] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=user.slice,mems_allowed=0,global_oom,task_memcg=/user.slice/user-1000.slice/user@1000.service/tmux-spawn-4577bb0e-6baa-4382-aadd-17e611ad469c.scope,task=sharehound,pid=1854,uid=1000
[Sun Dec  7 18:11:09 2025] Out of memory: Killed process 1854 (sharehound) total-vm:12158304kB, anon-rss:9259496kB, file-rss:328kB, shmem-rss:0kB, UID:1000 pgtables:20060kB oom_score_adj:0