TeaTimer memory usage

[Greyfox]

There are at least three separate threads on this subject now, see also

http://forums.spybot.info/showthread.php?t=30923
http://forums.spybot.info/showthread.php?t=30651

perhaps the forum moderator could look at combining them.

As it stands it appears that most XP users are experiencing what seems like unwarranted periods of sustained CPU activity when Teatimer is actived and also possible growth in the amount of memory used by Teatimer when the PC is run for long periods of time (leakage?).

Basically I think we are all now waiting for some official response on this subject.

[129260]

for most cases appears that the people that had this issue, at least most not all, just upgraded instead of a full clean install. I wonder if that will make a difference....

[wyrmrider]

Has Spybotsandra chimed in on this issue yet
she has been saying to just overwite
and she may be right
always
or
in most cases
or
never
too early to tell

I see the issue more often when the system has been upgraded more than once- but that is just an observation not a fact

[Greyfox]

for most cases appears that the people that had this issue, at least most not all, just upgraded instead of a full clean install. I wonder if that will make a difference....

Unfortunately that's not correct,

The characteristics that I have described in posts about Teatimer in the 1.6.0.30 version are from multiple PC's where a fully clean install was used. ie. Unimmunise, untick Teatimer and browser helper, untick host file and browser startup page lock, then uninstall old version. Manually delete folders in both program and document areas, run spybot small fix. Reboot, then install new version, download updates after install has finished.

If you have a look at Teatimer using Process Monitor, you will see what it is doing for the period of time after it is started, before CPU activity drops back - it is indeed working very hard possibly to establish a large registry reference and it does this every time it is started. The length of the busy period will vary with the speed and processing power of the CPU etc.

[bitman]

Greyfox,

Along with what Patrick has stated in the other thread and your own mention of CPU processing power, there are other influences. For example, someone else mentioned a maximum of 60-70% utilization which is probably caused by either limited memory and/or disk system response. While Patrick feels that most of these things will occur very quickly, he's forgetting how badly many real PC hardware systems actually perform, so his own test systems may not reflect the real experience of many users. This is especially true if the test systems operate as virtual machines with lots of RAM, since these usually operate much faster.

My own experience with TeaTimer has been an increasing degradation in startup performance on my old Windows 2000 based PII 400, even with the 512MB RAM it contains. However, that system is obviously processor starved, so TeaTimer is simply another element within a 10 minute plus startup sequence that also includes the real-time antivirus and other normal processes. There's no mystery here, it's simply more of the same as Patrick's post makes clear, the question is how to safely reduce this normal overhead while maintaining the security that TeaTimer process/registry monitoring adds.

What I think is more obvious, but less noticed by most is that a major decision about continued support for the Win9x/ME family will need to occur, if it hasn't already. The recent issues with the outdated 1.3 version of SS&D and the decision to discontinue detection update support for older versions force users to either upgrade or give up on Spybot S&D. The requirements of the likely solution to the TeaTimer startup issue will only stress the resources of these systems further. A clear decision needs to be made as to whether future support for the Win9x OS is even reasonable, since these OS are basically dead and would require significant changes to the operation of current versions of SS&D to continue real support.

Bitman

[tashi]

Greyfox,
Along with what Patrick has stated in the other thread

http://forums.spybot.info/showthread.php?t=30994

[129260]

Unfortunately that's not correct,

The characteristics that I have described in posts about Teatimer in the 1.6.0.30 version are from multiple PC's where a fully clean install was used. ie. Unimmunise, untick Teatimer and browser helper, untick host file and browser startup page lock, then uninstall old version. Manually delete folders in both program and document areas, run spybot small fix. Reboot, then install new version, download updates after install has finished.

If you have a look at Teatimer using Process Monitor, you will see what it is doing for the period of time after it is started, before CPU activity drops back - it is indeed working very hard possibly to establish a large registry reference and it does this every time it is started. The length of the busy period will vary with the speed and processing power of the CPU etc.

before i noticed ALOT of people have this problem, haha, but thanks for the correction.

[Coronamaker]

My experiences with the memory issue so far are that regardless of the system resources (RAM/CPU) or method of install (Upgrade or Clean Install) on all of the XP machines that I have put the 1.6 version on I have the same results. Upon completion of system start up, TeaTimer is typically using @ 27 MB which is better than the 46 MB most of my systems were usually at with the 1.5 version. FWIW I have not run into a creeping memory increase issue yet and I have been running this version through the Beta so maybe it is a conflict with other software.

If I use a memory cleaning program like Instant Memory Cleaner, that drops TeaTimer's memory use in Task Manager down to around 1.5-3.0 MB until the next time I launch an application at which time it hops back up to around 14.5 MB and stays there. Shutting down the application does not change the TeaTimer memory usage in Task Manager but re-running the memory cleaner again puts me back down to that same 1.5-3.0 MB roughly, and there it stays until I do something else. Not sure if there is a way to get TeaTimer to release that memory, but I am guessing it may result in an overall slowdown from having to release and then reload every time an application was launched. So now I just run a memclean after loading Windoze and let it use it's 14 MB.

[wyrmrider]

Interesting Corona
I'll try it with cacheman

[wyrmrider]

zerovlotage posts in the beta forum thread
responding here as his info in the other thread is valuable and do not wish to hijack it on a tangent

"Other programs with resident protection DO NOT take even longer. On systems where this problem exists, every other real-time solution works fine - only TT is showing spikes"

Counterspy shows exactly the same behaviour
spikes with at least two real time modules (active protection being one of them)
which can last a long time or forever
-with no way to stop except ctl>alt>del

and
if a scan is invoked "loading protection" can hang forever
i've let it run overnight without completion

happens after an update mostly