[Pulp-dev] Memory consumption on RPM sync

Dennis Kliban dkliban at redhat.com
Tue Nov 26 18:26:38 UTC 2019

The only difference between CentOS and RHEL repos is the Errata. So my
guess is that the memory consumption problem is introduced by how we keep
track of Errata and relationships to packages. Not sure how to fix it

On Tue, Nov 26, 2019 at 1:21 PM Fabricio Aguiar <fabricio.aguiar at redhat.com>

> I forgot to mention on my initial email, it was not an FYI,
> my intention in bringing these data was getting some discussion on what we
> should do to improve the performance and file some specific issues.
> Best regards,
> Fabricio Aguiar
> Software Engineer, Pulp Project
> Red Hat Brazil - Latam <https://www.redhat.com/>
> +55 11 999652368
> On Tue, Nov 26, 2019 at 3:03 PM Lubos Mjachky <lmjachky at redhat.com> wrote:
>> Great results!
>> I have tried to apply the same procedure like you, as I am currently
>> working on a similar issue (https://pulp.plan.io/issues/5689), but there
>> is one problem. I could not run performance tests with additional profilers
>> because my workstation run out of memory after a couple of minutes. I have
>> got only 16GB of RAM. Also, as you noticed, the synchronization takes a
>> huge amount of time. I think that we should not run those tests and
>> profilers on our personal machines.
>> For my tests, I used pulp3-source-fedora30 vagrant box, initially set
>>> with 4GB, then I increased to 16GB, and it was not enough. I only was able
>>> to sync RHEL 7 when I increased the memory up to 23GB.
>> What about using some allocated openstack instances? Do we have access to
>> some powerful machines as a team? I have managed to spin up some instances
>> but when I tried to run pulp services I always stumbled upon some issues.
> _______________________________________________
> Pulp-dev mailing list
> Pulp-dev at redhat.com
> https://www.redhat.com/mailman/listinfo/pulp-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/pulp-dev/attachments/20191126/dfb7c6e9/attachment.htm>

More information about the Pulp-dev mailing list