Recent Posts

Pages: [1] 2 3 ... 10
1
Process Lasso / Re: "3D" Colorization + Digest Report
« Last post by Jeremy Collake on June 21, 2021, 12:24:08 PM »
The graph additions, effectively adding new dimensions, would be cool. Certainly it could be accomplished, though it is a lot of work, and I am not sure where it fits into the overall agenda.

The next big planned enhanced for Process Lsaso is a tree listview so that process hierarchy can be shown, and collapsed. After that's completed, then I'll evaluate other big ticket items for Process Lasso.

In the end, I feel like you are describing a new product that's been on the radar for a long time, but that I have yet to dig into. Process analytics would be the primary function. It would be a large project, and there are some challenges, but I've always wanted to tackle it. Perhaps it will happen one day. In the meantime, I'll do what I can with Process Lasso ;).

Thanks for the feedback!
2
Process Lasso / Re: "3D" Colorization + Digest Report
« Last post by Jeremy Collake on June 19, 2021, 01:08:18 PM »
I certainly like the concept! I'll reply again after I've had time to chew on this.
3
Process Lasso / Re: "Restraints" column in main panel possible?
« Last post by Jeremy Collake on June 19, 2021, 01:04:02 PM »
Sure, I like the idea of a run count!

It would need to optionally differentiate instances based on their command lines.

I've now created an issue to track this potential addition.
4
Process Lasso / "3D" Colorization + Digest Report
« Last post by Coldblackice on June 18, 2021, 04:47:02 AM »
Would it be possible to add an element of colorizing the middle pane's text based on particular events or metrics?

For example, if a process hits X PB restraints/time, the text for that cell is changed to yellow. This could even go further with multiple stages of "severity", e.g.:

  • <3 = Default/White
  • >2 = Yellow
  • >5 = Red

Also possible (AND/OR) would be color-tinting (subtly) an entire process row if and when it hits a particular threshold, event, or alert, whether it's PBing, page faults, thread count, commit size, CPU time, etc.

The benefit of this is that, currently, PL presents data "two dimensionally": an XY-plane made up of column + row that triangulate to a single point of cell data. However, by adding color, not only is it visually stimulating to the brain, but it opens a "third dimension" of information -- and in the same amount of GUI space + layout!

(Warning: thesis-length spitballing ahead)

The question then would be "Well, what determines an appropriate threshold? When should 'white' become 'yellow'?" An initial simple possibility could be a basic flat comparison of average/overall stats amongst all (perhaps OS vs. non-OS) processes, e.g., "Of processes that have gotten PB'd in the past 24-hours, if the average is 2 PB's, any process >2 PB's gets its "PB Time" cell text bumped to yellow."

Now what would be really awesome for determining appropriate thresholds/metrics is if PL weighed the past X hours/days/weeks/months of data for that metric for each process (simpler than it probably sounds), and only if and when a process deviates beyond its daily/weekly/whateverly average for that metric, the cell (or row) is changed colors or bolded in order to stand out as anomalous, potentially warranting a deeper digging into, or at least keeping an eye on going forward.

This would be massively helpful in particular because there are so many processes running on an average computer, anywhere from sub ~100 to many hundreds of processes (currently 580 processes running on mine, I'm not proud to admit). It becomes an increasingly arduous  task for a person to know what processes are misbehaving unless it's starkly obvious, such as a program redlining 90% CPU or memory or I/O or whatever. But often times problematic processes or resource abuses don't manifest as easily, like a process that redlines for only short bursts of time but happens more spread apart, frequently enough to affect system performance but interspersed enough that the average user wouldn't catch it.



By adding color + threshold alerts, IMO it would evolve PL into not only a performance enhancer/enforcer, but upgrade PL into an omnipotent all-seeing sentinel, monitoring statistical log data and informing the user of problematic processes, problems, and/or resource inefficiencies that the average user wouldn't catch or even know about.

This would be massively beneficial in improving system performance because performance issues aren't always the fault of a singular process, but rather, two or more processes fueling a downstream tug-of-war over a resource or scheduling priority. Doing this would help make those problematic processes better stand out, helping the user to get to the bottom of the real source of the issue, e.g., which process is the primary problem/instigator, if it's looking like it's a driver or hardware issue that's affecting processes downstream, etc.



PL could have a button/function similar to "Insights" -- or even better, included in the Insights report -- that processes all the above data and gives the user a digest readout of potential problems in their system, similar to what Resplendence's "WhoCrashed" software does for BSODs/system dmp files.

Even being aptly technical, I still rely on a few tools like "WhoCrashed" that can digest technical data and spit out an "executive-level" summary, because we only have so much time in a day, and software that can do this to save us time or hassle is extremely beneficial. I can still pull out the tools to dissect a dump file if necessary, but I'd rather default to an automated digest analysis first, since most of the time a deeper-dive isn't necessary, like if I see it's due to a particularly buggy beta process/driver that I have to deal with for now.


Anyway, I realize this may be too beyond scope, and no problem if so, just spitballing some potential development ideas. Perhaps I could even make this on my own as an external third-party "mod" for PL, as an overlay or external report.
5
Process Lasso / Re: "Restraints" column in main panel possible?
« Last post by Coldblackice on June 18, 2021, 04:01:47 AM »
I was debating whether to make another post for this but don't want to spam, and it's in the same vein as this thread so I'll add it here:

Along with a historical Restraints/Insights column in the middle pane, another metric that what would be extremely useful is a column for "Run Count". I only just recently discovered a process that has been abusively running (called by another process) more than what should be necessary, however it's been occurring just barely spread apart enough that I haven't picked up on it during my regular brief glances through the bottom pane log to get a glance at what's been happening in my machine's background.

But if there was a "Run Count" column, I would've instantly seen "Whoa! This process has been called 300+ times in the past day! Let's find out why." Of course, run-count alone isn't necessarily indicative of a problem, but it can definitely be a useful piece of data in tracking down potential issues, be they downstream or upstream.
6
Awesome, sounds good! And no problem. I don't doubt it'd be a pickle to implement, I don't take for granted the work it must take to juggle all the varied development needs/priorities on your plate.

Anyway, looking forward to a hopeful/eventual release!
7
Process Lasso / Re: Auto Log Rotation
« Last post by Coldblackice on June 18, 2021, 02:04:12 AM »
Awesome, thanks. Please let us know if/when you add more!
8
Process Lasso / Re: Possible to remove columns from Actions log?
« Last post by Coldblackice on June 18, 2021, 02:03:23 AM »
Alrighty, no prob!
9
Process Lasso / Re: "Restraints" column in main panel possible?
« Last post by Coldblackice on June 18, 2021, 02:02:39 AM »
Note that the 'ProBalance Time' column of the All Processes list provides ProBalance restraint count and duration for specific process instances.

Therefore, I've assumed this feature request is to also show the historical ProBalance data across all instances of a process, as the Insights module does.

My bad, yes, this is what I was meaning, showing overall counts/time in sync with what Insights reports.
10
The highlights you see are ProBalance events, and have been there all along. They were improved a bit in v10 though.

I like the idea of showing more useful data about resource use spikes, or other process events. I also like the idea of making this configurable by the user. Time period indicators, for sure. There are some other neat ideas related to this that I won't delve into here.

This sort of graph improvement has long been on the radar, but it obviously takes some real design and labor to pull off, so hasn't made its way to the top of the planned feature list. I'll absolutely keep it in consideration. It may very well happen, but I don't want to commit to it right now.

Thanks for the feedback!
Pages: [1] 2 3 ... 10