lmacri wrote:
Apostolos wrote:
It appears that this issue is resolved.
-Fixed issue where high CPU and I/O usage during Idle (link)
Has anyone been able to confirm that the Norton v. 21.2.0.38 product update fixed the high CPU and disk I/O caused by SRTSP.SYS (Real-time AutoProtect) that starts after idletime Quick Scans? I haven't seen any noticeable improvement with NIS v. 20.5.0.28 on my 32-bit Vista Home Premium SP2 laptop.
Please note that the hallmark of this bug is a pop-up with the process name "No Activity" (i.e., not NIS.exe or ccSvcHst.exe) when you click inside the Performance Monitor graph, as shown in the screenshot below. After a re-boot, this activity will not appear during system idles until the first idletime Quick Scan has run. User elsewhere has posted instructions in message # 29 on how to reproduce the high activity caused by this bug.
Also note that the pop-up in the Performance Monitor graph will only say "No Activity" for periods when the NIS GUI was closed. If the GUI is open during data collection, the pop-up will typically display 1% NIS.exe (v. 21.x) or 1% ccSvcHst.exe (v. 20.x) but the process responsible for the remaining blue peaks of CPU activity (~ 15 - 20% of CPU on my machine) will not be identified, as shown in message # 36.
Symantec employee Basil posted here in slee2's thread NIS v21 High HD I/O that the high disk I/O that occurred in NIS 21.x when users enabled Advanced Events Monitoring in their Smart Firewall was fixed in NIS v. 21.2.0.38, but this bug fix was not mentioned at the bottom of the v. 21.2.0.38 product update announcement. I'm wondering now if the link in that announcement is pointing to the wrong thread.
-----------
MS Windows 32-bit Vista Home Premium SP2 * Firefox 29.0.1 * IE 9.0 * NIS 2013 v. 20.5.0.28
HP Pavilion dv6835ca, Intel Core2Duo CPU T5550 @ 1.83 GHz, 3.0 GB RAM, NVIDIA GeForce 8400M GS
Hi, Imacri. An idea:
Have you checked out the optional updates in Windows Update or Microsoft Update? There are a bunch of things in there that directly impact Power Management - along with updates which properly allow drivers to fully-and-completely-communicate all their capabilities with Windows itself.
For example:
On one of my machines here, I recently did a complete W7-X64 SP1 reinstall-from-scratch. (this was less than 2 months ago from the date of this post)
When I did the first set of Windows Updates on that machine (before I converted to Microsoft Update) - during one of the installs I left the machine running the update set and went to bed - because it takes an hour or more for that initial update set to be downloaded and installed on the particular machine I was updating.
Upon returning to the machine, it had automatically gone into hibernate mode after the inactivity timeout. When I hit a key on the keyboard to bring the machine back from hibernation, it spontaneously rebooted. When I checked the Reliability Monitor in the Action Center - it showed an item which said that Windows was not properly shut down.
I then checked the optional updates list. There was a specific update there which addressed this issue. I yet again reinstalled W7 from scratch - applied the optional update (KB2541014) - and was then able to successfully resume from sleep. With that in place, I was then able to properly and fully update an initial install of W7 using Windows Update - even if the machine had gone to sleep after the update process had completed.
The reason I mention this:
At that time, the standard set of Windows Update downloads for a W7-64bit SP1 initial-installation did not include this item by default - when it is clearly necessary (for some hardware configurations) for this update to be present before the system has even the remotest chance of allowing the user to walk away from something long-term (like the initial WU update cycle for a fresh install) - and have that process run to completion successfully without error.
Thus, I now have a healthy lack of respect for the accuracy of Microsoft's determination of what is "important" versus what is "recommended" versus what is "optional" - when it comes to Windows Update or Microsoft Update.
And on to the point of my post:
There are a bunch of updates in the "recommended" and "optional" categories which IMO provide important stability improvements for W7 operations. I am currently of the opinion there is nothing in there I consider unimportant.
All of them deserve installation - except for SilverLight, Bing, Microsoft Security Essentials, the Language Packs for Windows Ultimate which are not germane to the user - and possibly some of the Microsoft-recommended Driver Updates which need to be evaluated on an individual basis.
Please note that installation of some of these "optional" or "recommended" updates will trigger the installation of supplemental "important" updates. This again tends to support the idea that Microsoft's opinion regarding the "recommended" or "optional" categories is illusory at best. 
Normally - using the methodology recommended above - I end up going through about 4 or 5 update/reboot/recheck/download/reupdate cycles during an initial W7 installation - to get everything completely-up-to-date with all the what-I-now-consider-necessary "recommended" and "optional" updates in place.
My experience is this gives me a much more reliable, smoother and trouble-free Windows experience - especially when installing software that may have compatibility issues which are resolved by some of those "recommended" or "optional" updates - such as older versions of NIS, Microsoft Office, Norton Ghost, Symantec System Restore and more. Using the above procedure seems to be particularly important when it comes to Power Management issues with older Socket 775 motherboards (the cranky mobos here are the Asus P5Q-Deluxe and the Asus P5Q3-Deluxe boards).
Caveats, wrinkles and "gotchas":
I am not promising the above will solve your problem. Furthermore, Microsoft have subsequently revised their automatic detection for the W7 "important" patch list - such that manual installation of the KB2541014 update to solve the sleep/wake issue mentioned above is no longer required. However, in my experience this does not mitigate the need to install the "recommended" and "optional" updates that remain - even with the updated automatic detection routines in place.
I have not seen high CPU usage with SRTSP.SYS using the Windows Update "full update" methodology described above. I do remember having the problem with a much older W7 installation - which is one of the things which prompted me to do the "install all the recommended and optional updates" experiment in order to see if this solved the problem. It did not solve the Power Management issue I was trying to track down at the time - problems with improper installation of the Asus ATK0110 Power Management Driver - so I did not connect it with the issue you describe.
However, now that I have solved the Asus ATK0110 issue with the Asus P5Q-Deluxe and P5Q3-Deluxe motherboards, I am able to see linkages between certain Software issues and general Power Management issues - which I considered to be unrelated at the time. Thus, the above recommendation to at least perform the experiment.
As usual - I recommend a full Symantec System Restore 2013 (Ghost 15) or Acronis TrueImage backup - before performing the above experiment.
Hope this helps.
