[IMAGE: https://images.hive.blog/DQmauMDRNRVGJq5X92G6VukhKmDBTATNhbDUuREaxiuuryx/image.png]
Image of a spreadsheet showing the WAS requirement (explained later) not being met
Summary
GPUGrid is going to be greylisted due to it not sending out enough work units. It won't be included in the next superblock
What's Greylisting
Greylisting is essentially a temporary removal from the whitelist, but that doesn't require a poll to happen. If a project fixes whatever issue caused it to be greylisted, it will be added back to the whitelist (once the metrics are good) without a poll as well.
What's Causing GPUGrid to Have Low Amounts of Work?
It's not exactly clear why there are low work units available right now, but it looks like this is likely related to server issues going on with the project
What Triggers Greylisting
For GPUGrid, the Work Availability Score is what has triggered greylisting.
In general there are two criteria that can (directly) cause a project to be greylisted:
- Project Work Availability Score (WAS) is red
- Means that there are a low amount of credit recently given out compared to its past
- Means as that the seven day median new credit / the 40 day median credit is less than 0.1
- For GPUGrid this ratio currently¹ is
0.0100118245
- For GPUGrid this ratio currently¹ is
- Number of Zero Credit Days (ZCD) is greater than 7 out of the last 20 days
- Means that there are more than 7 days out of the last 20 that the project didn’t give out credits
- For GPUGrid this is currently¹ 5 out of the last 20 days
Other Notes
GPUGrid has gone on the greylist multiple times recently over the past few months. There may need to be more discussions around this project and whether it should stay on the whitelist
Footnotes
¹ As of 2021/07/28 (UTC)