100% disk utilization

I'm a stake technology specialist and I've gotten several complaints recently that clerk computers are unusably slow in the past couple of months.

For the machines I've looked at, task manager shows that disk utilization is pegged at 100% and often stays that way for 20-30 minutes (or more) after being powered on. The processes using disk are occasionally clearly an update, like some Edge updater. Other times it is just "system" or other vaguely named services. All of the computers have been replaced within the past 5 years according to schedule.

Has anyone else noticed an uptick of problems like this? Any suggestions? Re-imaging (per another thread here) seems drastic since there's not much on these computers in terms of data or any extra software.

The only idea I've had is to tell the wards to leave the computer on all the time so that updates can be applied and maintenance tasks run during the week instead of right when it needs to be used. But it seems odd that Windows would need so much high priority disk even for a machine that has only been on for a few hours a week.

What anti-virus is running on the machines? Sophos used to be used, but shouldn't be any longer. Yet due to a configuration error it tends to find it's way back on some clerk machines.

That AV will bring a machine to it's knees just doing it's own updates.

Currently, I think clerk machines are only supposed to be using Windows Defender.

Part of the issue with the older computers (like ours is) is that they have an old-slow rotary hard drive in them. They bog down extremely quickly even with very little usage.
I leave our stake clerk computer turned on permanently if I'm the last one using it, to try and let any updates happen out of hours.

Another possibility is a lack of adequate RAM can cause excess page swapping which can lead to HD "thrashing". Nowadays, 4GB is marginal to try to run Windows 10/11 on.

Thanks for the suggestions.

For one of the machines that has had this problem...

What anti-virus is running on the machines?

No sign of Sophos, though CrowsdStrike was using a fairly large chunk of RAM.

...they have an old-slow rotary hard drive in them.

I agree, it just seems like things have gotten worse recently for the same hardware.

Another possibility is a lack of adequate RAM...

This machine has 8GB, so not as bad as some.

I checked on the machine during the week, after having it left it on, and things do eventually settle down, even if it never seems to quite make it to something close to idle.

No experience with CrowsdStrike, but the symptoms sure sounds like an AV issue pushing a marginal machine over the edge. It downloads a new definition file and then does a scan.

You might want to consider programming the BIOS to power up an hour before the machine is needed. (And add a power policy to hibernate after 4 hours of no one comes in.)

I do have a weekly power up on the machines in our stake for early Wednesday so Patch Tuesday can process outside of operating hours.

I just run into this same issue recently. I was able to fix it by changing the virtual memory settings and defargmenting the computer. The computer is running very well again. Below is a good article showing how to do this.

https://softkeys.uk/blogs/blog/how-to-r ... 0then%20OK.

The current Lenovo computers have 5400rpm spinning disks which are outright terrible. Fortunately the Church's technology obsolescence management (hardware refresh) is purchasing Lenovos with SSD. I'd check with your FM group and if it's more than a year or so before you're up for refresh, I'd consider getting permission from your stake presidency to purchase NVMe SSDs for your computers, and then use cloning software to move the computer data from the disk drive to this NVMe. They're also crazy cheap, like $25 for a 256GB.

Thanks for all of the suggestions.

I've had people either leave the computers on or put them to sleep rather than powering them off all the time. I also turned on "Automatic Maintenance" in the Control Panel settings. Reports are that the computers are working much better. So either those steps helped, or there was something particularly heavy updating when trouble was reported.

I have our machines set in BIOS to start at something like 2AM on a Wednesday and several hours before they hibernate. The idea is to get all the MS updates taken care of overnight.