search

grayfallstown / Chia-Plot-Status

GUI Tool for beginners and experts to Monitor and Analyse Chia Plotting log files, show health and progress of running plots and estimated time to completion. No setup, configuration or installation of python or whatever required. Just install and enjoy.
https://grayfallstown.github.io/Chia-Plot-Status/
MIT License
183 stars 27 forks source link

Enhancement: Basic plot trends #61

Open revlisoft opened 3 years ago

revlisoft commented 3 years ago

Something to indicate how many finished plots in the past day. Maybe even a view of the past week to show how finished in each day.

Something like below to see if the changes made (either in hardware, threads, etc) improved overall number of plots being produced.

Finished plots

|Mon| Tue | Wed | Thurs | Fri | Sat | Sun |

| 14 | 12 | 14 | 13 | 15 | 11 | 11 | 12 |

The same could also be done for the other stats like Running, Concerning, Failed, Day avg for each phase.

grayfallstown commented 3 years ago

Some basic info will be in the next release.

grayfallstown commented 3 years ago

Released with 0.10.2

I wanted to make this visible with Charts but csharp seems to lack a good platform independend Chart-Library.

revlisoft commented 3 years ago

Released with 0.10.2

I wanted to make this visible with Charts but csharp seems to lack a good platform independend Chart-Library.

This is gorgeous and perfect! Slight sorting issue on the date but really amazing job! What does the numbers in the phases mean? Initially I thought there were averages but hours/mins doesn't seem to match. Then It kind of looks like count but shouldn't all the phases match finished or does it mean how many finished in that day for that specific phase? Not sure if count is as helpful but the average time to complete a phase is probably more useful to detect anomalies or improvements.

The screenshot below is from my laptop that connects to 2 machines via SMB to generate the logs. The 2 machines are built differently. I assume this just averages them all and treats them as 1 machine?

Here's a screenshot of mine!

image

grayfallstown commented 3 years ago

The numbers mean how many plotting processes entered a certain state at the given day. Basically, on Mai 27. 14 processed entered finished state, 12 entered Phase 1 (as in 12 were started on that day), 10 entered phase 2, and so on.

I assume this just averages them all and treats them as 1 machine?

It sums together all logs from all log folders. Could be separated to have one table for all, and one for each log folder to differentiate machines.

I am not sure if showing the avg time for each phase for a given day might be misleading. If you are HDD or SATA SSD plotting you get different times for each drive. On one day 2 SATA SSD and no HDD plotting plots finish phase 3: You would see the avg time of those SSDs. On another day 2 SATA Plot finish phase 3 as the day before, but this day the HDD finally finished phase 3 as well: You would see the avg time of those SSDs + of the HDD, which you did not get the day before. When plotting with 2 SATA SSDs and 1 HDD the user understands this. When mixing many different drives this becomes confusing.

A user might think his changes to the plotting process or system settings improved performance, but in reality it is just an HDD plot that finished yesterday, but not today which improves avg time for today. Worst case: The change the user made worsens performance but he thinks the opposite and he aborts all running plotting process so they benefit from the new change loosing a ton of time.