Terrell Russell

Confirmed/reproduced. Thanks @BartVanneste! ``` $ mkdir -p lots $ for X in {0001..5000}; do echo ${X} > lots/file${X}; done $ iput -r lots ``` Note: The bulk flag (`-b`) had...

However... the leak doesn't seem as large as reported above... # iRODS 4.2.8 on containerized Ubuntu16... an `iput` began with 200MB as a baseline: ``` 13015 /usr/sbin/irodsServer 00:04 197584 0.0...

very interesting. that helps narrow it down on our side as well.

I also suspected the number of resources present in memory may have an effect. I added 200 unixfilesystem resources during my tests above, and the leak speed did not change.

Not at this time. I just re-ran the above experiment with 4.2.11 and saw similar results as before. We clearly have a relatively small leak for each file, but when...

So the proposed fix is to just remove the `...` at the end? https://github.com/irods/irods_client_icommands/blob/0ec872ee34db6da61a37cde73fcf61166386df88/src/ifsck.cpp#L132

We're running through testing for 4.2.11 at the moment. If we find something else we need to fix, I'll pull this in as well.

Right. It's a lot easier to reason about results and return codes with a single input.

i notice that your first jargon row is not the smallest jargon pid.

is this now handled with 4.3.2.5-SNAPSHOT ?