« Perfcollect video series | Main | SQL Licensing and Virtualization: Lemons and Lemonade »

August 20, 2012


Graham Rhodes

This is a followup to my prior post. The problem I encountered downloading and running perfcollect4.1.exe was in fact due to my employer's web proxy, which trashed the download. It failed to download the last 14KB or so of the file. I connected to a different network, without the proxy, and downloaded again. The new download works fine.

Thank you for your attention!

Graham Rhodes

Graham Rhodes


I recently downloaded perfcollect4.1.exe (both interactive and unattended versions). There appears to be something corrupt with these self-extracting archive files that you are hosting. I found that the initial unzip never completed. After 15 or 20 *minutes* I killed via task manager, and in my user temp folder, perfcollect subfolder, I found the archive left a file called zip.exe that had grown to 75GB (yes, gigabytes) in size. (It was still growing when I killed the process.) I reproduced this problem on a 64-bit Windows 7 laptop, and a 64-bit Windows Server 2008 R2 server. The perfcollect4.1.exe interactive file size, as downloaded, is 950,272 bytes.

Are you aware of this issue? Can you test the download yourself, and perhaps upload a new file to see if that works?

Thank you!

Graham Rhodes

Paul Galjan

An excellent question, TC.

We've decided that perfcollect will remain a standalone tool to run locally on each machine/VM, primarily because the support issues associated with trying to collect remotely via WMI (any number of things can and do go wrong - firewalls, configuration problems, permissions, etc).

The good news is that we are considering adding a "framework" tool that can push perfcollect out to machines, execute it, and harvest the data. This framework could be extended work with other collectors, like ESX scripts, VNX/NAR data, etc.

This effort is being considered. To add your "vote", PM me and I'll put you in touch with the decision makers.

T.C. Ferguson

Does perfcollect currently handle (or do you have any plans to handle) automatic discovery/deployment/collection/etc?

The comments to this entry are closed.