ProjectMetricsUpdateTask causes high CPU usage

[navanin]

Good afternoon!
We are facing a high CPU load problem due to the ProjectMetricsUpdateTask. Apparently, this task is run every hour for each project, which causes a huge load on the processor, which leads to unavailability of the service and alerts from monitoring.

As I understand it, this is a kind of update of vulnerability data in existing projects. Can we optimize this task or disable it permanently?

DependencyTrack version: 4.11.0 (running by docker-compose)
OS version: Ubuntu 22.04

[navanin]

If I can provide any more useful information, please let me know and I will do it right away

[nscuro]

You can extend the cadence at which the task executes via Administration -> Configuration -> Task Scheduler -> Portfolio Metrics Update. Note that changing this setting will require a restart.

It's odd that it causes such excessive load though. Can you share more details about how many resources you assigned to the VM / Pod that runs DT?

[navanin]

Thank you for the answer!

I saw a mention of Portfolio Metrics Update in the project documentation, however, cadence is set to 3h, but the task still starts an hour after the container is launched.

Of course, I can share the VM configuration:
CPU - 6 vCPU 3.00Ghz
RAM - 24 GB

Is there anything else you might need to help me solve the problem? I would really appreciate!

[nscuro]

There is a hardcoded initial delay which might be one hour... After that initial delay, the cadence will take effect.

Is there anything else you might need to help me solve the problem?

What database are you using? Is the database running on a different VM than DT? If not, is it the database or DT itself that causes the spike in CPU usage?

Are you monitoring your DT application? Can you rule out a saturated database connection pool being the problem?

How many projects / components do you have in your portfolio? How long is the task running for?

[navanin]

PostgreSQL 16 is used, installed on a separate VM. The database VM shows a healthy load.

To be honest, I did not think about the connection pool, it is likely that this will solve our problem - I will check this information and come back with an update.

[navanin]

Nope, database is fine, the problem is not with the connection pool

[navanin]

We're getting LA 7 on CPU. This is really insane. Is there really no ways to disable or limit usage of this task?

[navanin]

@nscuro, can i completely disable this task?

[nscuro]

That is not supported.