Is there someway to record Task Managers info about CPU and memory usage to examine later? Or an equivalent tool?
-
Now in 2020, win10 Task Manager got [App history] tab showing seeming actual CPU time/cycle usage (whereas procexp CPU time column seems more like time since process start) – was Feb 07 '21 at 05:31
8 Answers
Windows Performance Monitor (perfmon) should do the job for you; you can configure it to log to a file, so just enable the counters you need and it'll log as much as you want.
![](../../users/profiles/6965.webp)
- 3,700
- 1
- 23
- 29
-
14Any chance you want to write up a more complete answer? Some steps to follow? – Buh Buh Oct 28 '16 at 13:33
-
See answer below by @Matt (and upvote; he deserves it) – Mawg says reinstate Monica May 19 '19 at 06:49
- Start > Run > perfmon
- Hit the plus sign next to Performance Logs and Alerts
- Right click System Log and select properties.
- Adjust the sampling times to whatever you like
- When you are done, hit OK then the plus sign at the top of the window.
This page gives a pretty good step-by-step with screenshots
![](../../users/profiles/407822.webp)
- 123
- 8
![](../../users/profiles/10288.webp)
- 1,112
- 1
- 8
- 18
-
5
-
5
-
1"Hit the plus sign next to Performance Logs and Alerts" Where is that? – endolith Jun 09 '21 at 23:33
-
To collect the required Perfmon log data in Windows 7, Windows Server 2008 R2, Windows 8, Windows 2012 and Windows Vista: Click Start > Run, enter perfmon.exe, and click OK. Go to the User Defined folder. Click New > Data Collector Set. Enter a name for your Data Collector Set. https://kb.vmware.com/s/article/2010970 – endolith Jun 09 '21 at 23:39
I use PolyMon to monitor these kinds of things. You can define various "alert" conditions about which you'd like to receive notification, so I get notified if one of my servers is having a problem.
But it also stores all these results in a long term database, so I can look back at the memory usage of server "X" and see it's memory usage trends over the last N days/weeks/months/years.
![](../../users/profiles/4810.webp)
- 597
- 2
- 8
Opensource solution Nagios does this also:
You may consider using a monitoring solution like zabbix, zenoss to collect and analyse your perf counters.
![](../../users/profiles/8132.webp)
- 5,026
- 1
- 25
- 31
Cacti is a great webv based graphing programme. It can graph CPU, memory, disk space, etc, etc. It can also be extended with your own plugins.
![](../../users/profiles/8950.webp)
- 30,211
- 62
- 184
- 246