From SPC14: “Troubleshoot Search” session (spc375)

Wow! I wanted to send a huge thank you to the >500 folks that attended our session today and for the very gracious feedback. Here is a link to our session on Channel9:

How to manage and troubleshoot Search: A practical guide (SPC375)


And as promised, here's the PowerShell that I build for filtering the ULS by correlation Id and having it automatically open in ULS viewer. To run it, simply use:

.\getULSbyCorrelation.ps1 "f41cb190-945e-458e-b924-77ec2fd066d4"  


$pathToULSViewer = "D:\Toolbox\UlsViewer.exe"

if ($args[0] -ne $null) {

    if ($args[1] -ne $null) {

        $outputPath = $args[1]

    } else {

        $outputPath = Join-Path $Pwd.Path $("byCorrelation_" + $args[0] + ".log")


    Merge-SPLogFile -Path $outputPath -overwrite -Correlation $args[0] | Out-Null 

    if (Test-Path $outputPath) { Invoke-Expression "$pathToULSViewer $outputPath" }

    else { Write-Warning ("=== Found No Matching Events for this Correlation ==="}

} else { Write-Warning ("=== No CorrelationId Provided ===")  


And optionally, specify a target path for this run using:

.\getULSbyCorrelation.ps1 "f41cb190-945e-458e-b924-77ec2fd066d4" $fullPathAndFileNameForOutput


Very much appreciated to all - this was a REALLY rewarding experience!

--Brian and Jon

Comments (4)
  1. Joe Starr ( says:

    Brian, can you post a link to the script for setting up that search server?

  2. bspender says:

    Hey Joe.. .good to meet you earlier.

    The script used in that demo for provisioning an SSA was from an earlier post here:

    Deploying SharePoint 2013 SSA Across Multiple Servers Using PowerShell…/deploy-sharepoint-2013-ssa-across-multiple-servers-using-powershell.aspx

  3. Luis Valencia says:

    Hello Brian

    I didnt find how to contact you directly so I will post my question here.  After looking at many spc 14 I just watched SPC375, and I think you are the right person to ask.

    We have a sharepoint farm with 2 wfe and 2 app servers, we have 2 web apps, one for our intranet (35gb of data), and another web app for our client documents system, which is 20.000 site collection spread over 5 content db, max 300gb.

    We created a content source for each.

    A full crawl takes about 24 hours at a dps of 10, and incremental crawls takes about 2 hours  with a dps of 2-4.

    When I checked the crawl health reports I see a repository latency between 3000-4000ms, CTS latency betwee 1000-2000ms and SQL latency is low, (0-30ms).

    I havent still found any information about how to troubleshoot this, I watched your session and maybe I can track a document through through ULS, I am not sure if this helps because the documents are crawled, crawler is not stuck.

    The intranet content source makes a full crawl in 15 min, and incremental crawls in 1 minute, so the problem  its only with our client document system's content source.

    Would you mind pointing me in the right direction? how to troubleshoot repository and cts latency?

    Thanks a lot

  4. Luis Esteban Valencia M says:

    Hello Brian, after researching a lot, and seeing your SPC375 with Jon I think you are the right person for my quesiton.

    We have a farm with 2 web applications, one hosts the intranet with 35gb of data, full crawl takes 15 min, incremental takes 1 min or 2.  The 2nd web application is the document management system with 20000 site collections, and over 300gb of data in 5 content databases.  On this one a full crawl takes 24 hours and incremental cralws takes 2 hours.  On each incremental crawl succeses are about 30,000, when we check the succeses its mainly Allitems.aspx and the homepage of each site collection, I know this doesnt change a lot.

    The main problem is crawl health, in the reports we can see repository latency between 3000 and 4000, and CTS latency between 1000-2000, sql latency is very low between 0-30.

    I wonder if you can guide me in the right direction to troubleshoot this latencies as they are too big.  I know I could ignore AllItems.aspx from being crawled but that will only reduce the nr of succeses, it wont reduce the main problem, latency.

    After seeing the SPC video, I am not sure if tracking a document through the uls logs will help me, because my crawler is not stuck, its only too slow.  Our dps in incremental crawl is 3 or 4, on full crawls its 9 or 10.

    Thanks a lot in advance.

Comments are closed.

Skip to main content