High Load on IO Subsystem Running FullCrawl

[ultrator]

Hi

If i run a FullCrawl on my SharePoint 2019 on Premises it generates high load on the IO Subsystem and brings the AlwaysOn Cluster wich holds the SharePoint Databases to FailOver oder Crash.

We have 3 Crawl Servers and 3 FrontEnd Servers for Query Processing and the index.

Can i optimize the search architecture or slow down the crawl to balance the load?

Here you see the last run of the FullCrawl:

I am at a loss. Where can I begin?

Faster IO Subsystem is ATM not a option...

Regards

Ueli

[#2 · Accepted solution]

You can use Crawler Impact Rules to reduce the load. This increases the time required to crawl all content but reduces I/O on SQL.

https://docs.microsoft.com/en-us/sharepoint/search/best-practices-for-crawling#using-crawler-impact-...

You should also look at the configuration of SQL Server. Make sure the logs are on different physical disks than the main database.

[#3]

Hi @Steve Knutson 

Thanks, that's what i was searching for.

I'll try the crawler impact rule.

Best regards

Ueli