Server Monitor

[cju-mt]

The quick shout out Joseph, xyOps is incredible. Just wanted to get that out there before I mention the problem I'm running into.

Server monitors for Window Satellites do not have "Quick Look — Last Minute" like the linux satellites do.

Live monitoring as well as "Last Hour" (for disk usage, Disk Read, Disk Write, CPU I/O, Disk I/O, Open files) doesn't appear to work.
Disk Usage always says 0%
Open FIles 0
CPU load 0
etc...

Is there anything that needs to be added to the satellites to make this work? Python
It would be really cool to be able to get this working.

Thanks,
Dave

[jhuckaby]

Hey, thank you for the kind words ❤️

Server monitors for Window Satellites do not have "Quick Look — Last Minute" like the linux satellites do.

Yeah, that's a known issue. Unfortunately, its very difficult to get real-time CPU, memory, disk and network utilization numbers on Windows.

I had it working in an early development version, but it caused a huge amount of CPU usage just to read the metrics, and it took upwards of 5 seconds PER READ. It was just ridiculous, so I had to disable it.

I have it on my TODO list to revisit this after I launch v1.0. I'm hoping the upcoming big v6 rewrite of systeminformation (the library I am using to gather metrics) has better support for Windows.

Live monitoring as well as "Last Hour" (for disk usage, Disk Read, Disk Write, CPU I/O, Disk I/O, Open files) doesn't appear to work.

Okay, this is not good. I've tested this thoroughly across a bunch of different Windows machines. Can you tell me more information about your setup? Specifically, exactly what Windows version you are running, what version of xyOps, and what version of xySat (xyOps Satellite)? Was xySat installed normally and is it running as administrator?

The CPU / disk / IO numbers come directly from systeminformation, so it may be that it has some kind of compatibility issue with your machine.

[jhuckaby]

A few things to note about Windows:

• CPU "Load Average" is a Linux invention -- windows doesn't have an equivalent, so this is always zero. If all your servers are windows, you can completely disable this monitor.
• CPU "I/O Wait" is also a Linux thing -- Windows doesn't have this either, so it's always zero.
• The "Open Files" monitor is a Linux-only Plugin, provided an example to show how you can add custom monitors. It will always be zero on Windows.

I do see that Disk read/write are stuck at zero -- I am looking into this, but it seems like the systeminformation library may be the issue 😞

[jhuckaby]

Ugh, this systeminformation issue has been open since 2018: sebhildebrandt/systeminformation#138

Maybe I can try some of the workarounds suggested in the thread.

As you can see there are quite a few open issues with Windows:

https://github.com/sebhildebrandt/systeminformation/issues?q=is%3Aissue%20state%3Aopen%20windows

I feel for the author -- it's a really tricky platform to get accurate (and fast) metrics on.

[nickdollimount]

cough cough Windows sucks cough

[talder]

Oh come on, Windows Vista was just awesome, admit it! 🤣

The server editions are fine, especially the 2012R2 and 2022. Hell, we had more trouble with our Solaris machines than our Windows and Linux servers, can you imagine.

For desktop this was XP, 7 and 10. What they do now with 11, god may know, but I moved from Windows to Linux and now on Mac. All have flaws, but the biggest reason for me to go to Mac was software compatibility as a system engineer and the abomination that Windows has become with is tracking/pop-up/bloatware/update stuff. After using Linux for over 3 decades, I'm pretty sure Linux desktop will never be a thing. 😞

[cju-mt]

Problem solved. Created a new monitor for windows satellites.

I am using mounts.C_.use for the monitor and alert.

I created what I call xyOps Honeycomb. It uses a series of API calls to iterate all satellites each represented by a hexagon. The hexagon will flash different colors based on the severity of the any alerts for that satellite. Allows me to have a monitor screen up that monitors all servers and computers and insures they are up and operating within alert limits. Even gives an audible alert. The possibilities are absolutely limitless.

I am super jazzed with xyOps.

Thank you!!!!!!!!!!!!!!!!!!!!!!!!!! 😊😊😊

[jhuckaby]

OMG I WANT TO SEE THE HEXAGON! SHOW US THE HEXAGON!

Hexagon is the bestagon, after all.

😉

(Also, congrats on solving the problem! Makes sense, because the default disk usage monitor is looking at the "root" (/) mount, which is a Linux thing).

[cju-mt]

Here you go. Credit goes to
https://github.com/hasielhassan

• Hex grid layout: HoneyCombLayoutJs (MIT)
• Inspired by CheckMK

[jhuckaby]

WOW! That is really cool! Love it 😃

Thanks for sharing!

[ovizii]

* Hex grid layout: [HoneyCombLayoutJs](https://github.com/hasielhassan/HoneyCombLayoutJs) (MIT)

Do you mind sharing a few words on how you combined that with xyOps?

[cju-mt]

I am having an issue with alerts. They trigger when you test them but do not remain as persistent alerts. After I believe 24 hours they get cleared, never to appear again. If I clone the alert they will again remain persistent, for how long, we will see as I just cloned it.

Hopefully this is just a setting :)

Thanks

[jhuckaby]

That's really strange! Let me do a test here, with an alert that is always triggered...

Alert active at 2:46 PM Pacific. Will check back tomorrow...

[jhuckaby]

Okay, I was able to reproduce this. It happens when a remote server disconnects with active alerts, then a minute goes by and the primary server expires the alerts, and then the same server reconnects. It doesn't "resume" the alert correctly, and instead puts the alert into a weird phantom state where it goes invisible.

Working on this ASAP!

[cju-mt]

I am glad you found it or at least are hot on the trail. Did I tell you how much I love xyOps :-)

…

-Dave On Feb 22, 2026, at 8:11 PM, Joseph Huckaby ***@***.***> wrote:  Okay, I was able to reproduce this. It happens when a remote server disconnects with active alerts, then a minute goes by and the primary server expires the alerts, and then the same server reconnects. It doesn't "resume" the alert correctly, and instead puts the alert into a weird phantom state where it goes invisible. Working on this ASAP! — Reply to this email directly, view it on GitHub<#138 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AC2ANZZUBWHMR2BE4XMXOD34NJVU3AVCNFSM6AAAAACVCVLMO2VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKOBZGI3DONA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[jhuckaby]

Thank you! Very glad you are enjoying it.

Fixed in v1.0.8.