Just look and in Dez 19 my host receive a set of few weird expiration time CPU WU

An example of one: https://setiathome.berkeley.edu/result.php?resultid=6245493793

Received: 19 Dec 2017, 12:29:16 UTC
Expired: 19 Dec 2017, 23:41:11 UTC

Less than 12 Hrs to expire after they where received.

Something breaks on the scheduler? Or is in my host? Do I need to do anything to avoid this?

---

Sometimes the server just screws up. I have seen deadlines of just a few minutes, but that this is rare to occur.

Sometimes the server just screws up. I have seen deadlines of just a few minutes, but that this is rare to occur.

I always thought it was when the server screws up and sends a Arecibo VLAR to a Nvidia GPU. Those with a Sent time and Deadline time within a few minutes of each other which results in the expected task not completed by deadline.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

Just look and in Dez 19 my host receive a set of few weird expiration time CPU WU

An example of one: https://setiathome.berkeley.edu/result.php?resultid=6245493793

Received: 19 Dec 2017, 12:29:16 UTC
Expired: 19 Dec 2017, 23:41:11 UTC

Less than 12 Hrs to expire after they where received.

Something breaks on the scheduler? Or is in my host? Do I need to do anything to avoid this?

It looks like those were sent about a half hour before the Tuesday outage, but didn't finish downloading before the outage started, for some reason. When the servers came back up 11 hours later, they immediately decided that they'd waited long enough for the downloads to complete and just cut them off. Your Event Log might show why they didn't manage to download in that half hour before the outage, but it may not be anything you could have done anything about.

I always thought it was when the server screws up and sends a Arecibo VLAR to a Nvidia GPU.

I've only seen that happen when I'm recovering "ghosts", but then they get marked as "Abandoned" rather than "Timed out".

I always thought it was when the server screws up and sends a Arecibo VLAR to a Nvidia GPU.

I've only seen that happen when I'm recovering "ghosts", but then they get marked as "Abandoned" rather than "Timed out".

Yes, I see that as the most common scenario. But I have also seen these tasks downloaded in a normal scenario in the log and then only a few minutes later reported as timed out. Rather rare but I have seen it a couple of times on all machines over the years. I just put it down to the servers having a brain fart.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours