Microsoft tweaks Windows 10 indexing system with May 2020 update

Shawn Knight

Posts: 15,303   +193
Staff member
Recap: Years ago, it wasn’t uncommon for Windows tweak guides to suggest disabling indexing altogether to boost system performance. This slowed down searches but kept the OS from unnecessarily spinning up the hard drive at the most inopportune time and was thus deemed a worthy trade-off for many.

Microsoft’s upcoming May 2020 Update for Windows 10 will reportedly include a tweaked indexing system that could boost overall system performance with certain configurations.

Indexing in Windows involves the operating system examining the files on your computer and cataloging their information. The next time you search your system for a specific file or keyword, it looks to the index to help find results faster.

The problem is that indexing can be taxing on your hardware and lead to heightened disk activity, especially if you are using a mechanical hard drive. This is where the new update comes in.

As Windows Latest highlights, Microsoft has fine-tuned the indexer in the latest update so it should have less of a performance impact. According to the publication, Windows will now throttle or even completely halt indexing while you are doing other activities like transferring files. The site said it was able to notice a difference in multiple tests they ran on real-world hardware and virtual machines.

Worth noting is the fact that this tweak will make a much bigger difference on HDD-based systems than it will on those running a solid state drive. Flash-based drives are typically fast enough that indexing may not be noticeable to begin with.

Masthead credit: denniro

Permalink to story.

 
I believe Windows 10 has a bigger indexing problem. But, it doesn't impact everyone. When I work on customer computers, I pull the HD/SSD and connect them to one of two workstations, so I can image the drive and manually remove any nasties found. Windows 10 takes from one to twenty minutes to allow me to see the User account files, before they show in File Explorer. If I put the same drive, or restored image of it, on my Windows 7 machine, it shows the User account files immediately. This issue, which again does not impact just about everyone else, is terribly annoying. So, I pay for 0patch for the Windows 7 machine and will remove Windows 10 and replace it with Linux Mint, if this continues after the next couple of updates. Productivity is impacted. I have had Windows 10 Pro and Enterprise LTSB on Workstation 1 and I'm not happy with it's "you don't have permission, click here to give yourself permission" nonsense also. That just phony security. I really like everything else about Windows 10, believe it or not.

That's my two cents on Windows 10 search. I use my Windows 7 machine almost all the time, and 10 seldom these days. I know Windows 7 support will be non-existent in three years. I'll cross that bridge when I come to it.
 
4tb hdd is around 100 bucks. 4tb ssd is around 500. Anyone saying hdd's are completely irrelevant on a desktop pc in 2020 either don't have much data to store or can afford to pay 4-5x as much per tb.
 
4tb hdd is around 100 bucks. 4tb ssd is around 500. Anyone saying hdd's are completely irrelevant on a desktop pc in 2020 either don't have much data to store or can afford to pay 4-5x as much per tb.
HDDs will be relevant for quite a long time cause price and reliability.
 
SSD's prices might've gone down quite a bit, but HDD storage is still dirt cheap in comparison (10x cheaper last I checked).
Using HDD's as storage drives will be a thing for a while yet, so speed ups are always welcome...

how does indexing helps you with storage HDD exactly? it's rhetorical question.
 
If they ever can get dual independent heads working on the drives, then much of the speed problems can be solved.
 
how does indexing helps you with storage HDD exactly? it's rhetorical question.
I am assuming you may not be familiar with file systems, if I assumed wrongly, I apologize ahead of time. In the most simplistic view, if "files" are stored in a "sequential" manner in the order they are created, you would spend a lot of time scanning the hard drive when you went to retrieve/update it. By using indexing, you can create a secondary file (index record), inserting the filename and the address as a "pointer" to the "stored file" alphabetically. This secondary file would have multiple entries and require much less hard drive space and be quickly searched to find the hard drive address of a file. As the file system grows, you would create more index records, which, in turn would have an index of the index records and this would develop into a tree structure of indexes. In this simplistic view, whenever you need to access a file, you read the index records, scan it in storage for the filename and address and then retrieve the file. If lowest level index consisted of many records, then they would be indexed using either the first or last entry in each index below with the address of the lower level index. So to access a given file, you would start at top index (maybe kept in storage to eliminate a hard drive read) and follow the index structure to locate the actual file you want. So depending on the number of files, you can locate the desired file in a few reads vs scanning file after file looking for it. The directory (folder) structure of the Windows (and Unix/Linux) is basically an index structure. You have the "root" directory with the various sub-directories under it.

As I have indicated, this is a very simplistic view and depends on how the file system structure is designed. For instance, I might have a set of indexes, one for the operating system files and another set for the user files. This reduces the amount of index reads required to locate the system file or the user file. The real old hard drives had actuator arms to move the read/write heads over the rotating "platter(s)" on which the files are stored. The newer hard drives use magnetic coils to position the read/write heads faster. One can also use a "randomizing" technique to locate files based on using the filename (or other identifying factors) to generate an hard drive address, but that is a different topic. In any case, spinning platters, actuator arms, make hard drives "quite slow" compared to SSDs. There is an interesting article on hard drives, see: https://www.explainthatstuff.com/harddrive.html . And, of course, you can check Wikipedia too.

As to the type of indexing discussed in this article, Microsoft has taken it a step further and scans the files in the file system for relevant types of information and sorts them into catalogs of indexes. So when you want to search for a certain word/phrase/filename/etc., it uses those indexes to locate the files which contain the search word(s) and displays the list of items it finds. So if you are looking for a certain document in your Documents directory that contains a specific topic, rather than scanning every document in the directory, it can find the information in the indexing catalogs. For more information, see: https://docs.microsoft.com/en-us/previous-versions//ms689718(v=vs.85)?redirectedfrom=MSDN . Another article you may find of interest is: https://community.adobe.com/t5/prem...-and-how-can-it-affect-me/td-p/2797052?page=1 .

 
Back