WARNING - Looking up qualities on plot takes more than 5 sec

I’m getting many warnings of plot look up qualities times higher than 5 secs.
This is new. I haven’t got that in the past and I’m not getting any rewards for over a week.
I have 330 TiB and everything seems to work fine, so expectation to win one block every 2 days now, and a week ago one block every day.

Could the reason be the look up time? And Is there any way to fix this?

The big one is harvester time. Harvester checks must return in under 30 seconds, or you will never win awards.

haven’t seen harvester time mentioned in the logs, guess that’s good

I had this on a couple older drives that once in a while have a very slow response. I took one out of the farm for now. But the bothersome thing was how it hung up the log output. I only noticed this happen because I ran “chia farm summary” and it hung there for 15-20 seconds and no debug.log output either. Then, suddenly a huge log dump happend and the summary showed. At the top was that error. It was almost as if chia was hung up. So, yeah, took that drive right out! Haven’t seen the error since. I was thinking chia was more asynchronous though.

Harvester warnings:

2021-05-11T12:59:42.681 harvester chia.harvester.harvester: WARNING Looking up qualities on N:\final\plot-k32-2021-05-04-08-00-e577a481...................a485.plot took: 10.492086410522461. This should be below 5 seconds to minimize risk of losing rewards.

It says times should be under 5 sec.

5 vs 30 seconds has always confused me. Before the warning addition in logs, people kept saying under 30 seconds. But then they put the warning down at 5 seconds? That’s a huge gap.

30 seconds is a hard limit. If you don’t reply in 30 seconds, you will never get the reward.

5 seconds is a soft limit. If you are above 5 seconds regularly, you should look at your setup and find your congestion point.

2 Likes

What might be a possible congestion point?

drive I/O. Usually if you are copying a new plot to a HDD the response time for the look up goes up.

But I’m getting all these warnings on drives that are already filled up with plots.
The times used to be less than 5sec. Something changed in the last 10 days - may the software, maybe my configuration

Any other idea for congestion?

ahhh. i am getting those a couple times a day. but they are just beyond 5 seconds. typically like 5.234762534786234 seconds. think that’s safe?

So, anyone have an idea how to lower these harvester times?

If you actually find a proof and win do you know how many filters it passes and what the total lookup time is?

Unfortunately no, but the chia team member said -n 30 was fairly close to that.

1 Like

It looks like you are using windows. I was getting these errors before. I haven’t had them since I made the following changes:

1 - Go to “Power & Sleep Settings”
2 - Click “Additional power settings” on the right panel.
3 - Expand “Show additional plans”.
4 - Select “High performance”.
5 - Select “Change Plan Settings”
6 - Select “Change advanced power settings”
7 - Under Hard disk → Turn off hard disk after, set this to 0
8 - Under USB Settings → “USB selective suspend settings”, set this to Disabled.

7 Likes