Requesting AV Disk I/O Experiment

Discussion in 'other anti-virus software' started by Brandonn2010, Dec 2, 2012.

Thread Status:
Not open for further replies.
  1. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Well my last thread about disk space used by AVs turned out to be useless, but what if someone with a VM or just test machine could measure the disk I/O of all the main AVs, and we can list them here to see which have the highest and lowest, as this seems to affect performance the most rather than RAM usage. It would need to be the same machine each test so there wouldn't be other variables.
     
  2. clocks

    clocks Registered Member

    Joined:
    Aug 25, 2007
    Posts:
    2,559
    I nominate you to take this project on. :thumb:
     
  3. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    I have neither the time nor the know-how of what to do. I'm not even sure how exactly to measure the disk I/Os.
     
  4. Macstorm

    Macstorm Registered Member

    Joined:
    Mar 7, 2005
    Posts:
    2,531
    Location:
    Sneffels volcano
    Says who? It's an alive and kicking open thread yet ;)
     
  5. DBone

    DBone Registered Member

    Joined:
    Nov 24, 2010
    Posts:
    1,041
    Location:
    SoCal USA
    Hey Brandon, a little off topic, but have you thought about ditching your AV and just running with AppGuard? That's the way I run, and all of your I/O worries will be a thing of the past. :thumb:
     
  6. zfactor

    zfactor Registered Member

    Joined:
    Mar 10, 2005
    Posts:
    6,012
    Location:
    on my zx10-r
    i have plenty of machines. what info would you want from each av? and how would you want it to be tested like for how long since disk i/o increases over time, and when testing ram what would you want to be shown like under idle then while scanning etc??

    have plenty of machines i can test stuff on here and def have one i dont use so they all would be on the same machine would have to test one at a time. i also would not use a vm i would simply create a baseline image and restore between each av installation i feel its better than using a vm personally for testing.
     
  7. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Actually I have gone with just AppGuard, but I felt like something was missing. I reinstalled Avast! and after it got its cache back, my computer is no faster or slower than when I was AV-free.

    I'm not too worried about I/O for me since I have a fast PC, and I know Avast! is fast, I just want to know which are heavier/lighter by default.
     
  8. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    I'm not sure what to test which is one of the reasons I was asking everyone else. I have just heard things such as Avast! having low disk usage, Panda cloud high, etc.

    AVs I would want tested would be:

    Avast! Free
    AVG Free
    Avira Free
    Panda Cloud Free
    Roboscan Free
    ZoneAlarm Free AV + FW

    then probably

    BitDefender AV
    Kaspersky AV
    Norton 360 since so many people use the free one from Comcast

    Activities to measure disk I/O during would be web browsing since everyone uses a PC for that, possibly playing video games, and copying files.
     
  9. KelvinW4

    KelvinW4 Registered Member

    Joined:
    Oct 11, 2011
    Posts:
    1,199
    Location:
    Los Angeles, California
    It would be nice to see Zonealarm, Avast, and Panda.
    Maybe MSE too?
     
  10. zfactor

    zfactor Registered Member

    Joined:
    Mar 10, 2005
    Posts:
    6,012
    Location:
    on my zx10-r
    what we would need is a set base. meaning a set standard. like a specific list of web pages to be opened and how many etc.

    as well as a list which programs to be used etc. i could prob test at least one or two a day depending on how long of a test period required.

    though this may be for nothing because most of the time they are going to remove any self testing here at wilders.
     
  11. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    It should be fine. It's not a malware-testing test, it's a performance test, and if done in a consistent controlled way, it should be fine.
     
  12. lodore

    lodore Registered Member

    Joined:
    Jun 22, 2006
    Posts:
    9,006
    I just remembered that passmark already produce a report every so often doing all sorts of benchmarks with antivirus software including installation size and disc I/O
    http://www.passmark.com/benchmark-reports/
     
  13. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Is that sponsored by Symantec? If not it sure seems like it..
     
  14. jo3blac1

    jo3blac1 Registered Member

    Joined:
    Sep 15, 2012
    Posts:
    739
    Location:
    U.S.
    +1 to this idea. RAM usage and disk space usage is largely irrelevant. So this would actually compare how heavy each AV is.
     
  15. lodore

    lodore Registered Member

    Joined:
    Jun 22, 2006
    Posts:
    9,006
    If a product can protect as well but use less resources and less disc space then that would be a reason to go for it.
     
  16. Notok

    Notok Registered Member

    Joined:
    May 28, 2004
    Posts:
    2,969
    Location:
    Portland, OR (USA)
    Passmark also has an app launch timer: http://www.passmark.com/products/apptimer.htm

    Just keep in mind that after the first launch the app may be cached in memory, and so it'll launch faster the second time. AV software also tends to only scan files fully the first time until it changes. With that said, I did see some small differences when enabling and disabling realtime protection and relaunching the same apps.
     
  17. Kees1958

    Kees1958 Registered Member

    Joined:
    Jul 8, 2006
    Posts:
    5,857
    Well, being a security minimalist, using a desktop with a simple E5200 dual core and reasonable fast 500GB HD now (since old one crashed), I have not tested all, just the usual freebies, like Avast, AVG, Avira, MSE and Panda. Tested a year ago, so today's result could show different results, old hard disk at that time averaged 70 MB/s throughput. Tested on real system (no VM).

    - fastest Avast in terms of second on impact on reads & writes and with lowest overall disk I/O
    - Avira second (the other way around, but Avast was very close in read&write delays hence Avira came second)
    - AVG (disabling Linkscanner, had serious impact on CPU %)
    - MSE
    - Panda (even after full scan)

    It is a lot of work, goodluck with your request :thumb:
     
  18. steve1955

    steve1955 Registered Member

    Joined:
    Feb 7, 2004
    Posts:
    1,384
    Location:
    Sunny(in my dreams)Manchester,England
    what difference would that make to objective figures?
     
  19. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Because I would question the objectiveness of the figures, especially since Norton did the best, despite its reputation for sluggishness, even with the post-2009 versions.

    Regardless, I would still like this test done, as out of free AVs, only Avast! Free was on there, and I was surprised how poorly it did, given its AV-C results.

    Also, I would add Kingsoft Cloud AV to the list.
     
  20. m0unds

    m0unds Guest

    for the results to be meaningful, you'd really need to establish how to measure i/o (e.g. not using the task manager read/write bytes/ops column since this includes /all/ types of io, and not just disk stuff.)
     
  21. zfactor

    zfactor Registered Member

    Joined:
    Mar 10, 2005
    Posts:
    6,012
    Location:
    on my zx10-r
    thats kinda my point is we would need to establish a test method and baselines to get accurate testing
     
  22. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Somewhat off-topic but what about firewalls with HIPS? Do they have much disk I/O? Has anyone measured the heaviness of COMODO , Online Armor, Private FW, Outpost?
     
  23. m0unds

    m0unds Guest

    i nearly always equate the use of the word "heaviness" within the context of anything on wilders as anecdotal in nature. i'm sure somebody somewhere has measured system impact of the products in question, but i don't think i've ever actually seen any such thing (specific data w/results and methodology anyone could replicate themselves) on wilders.
     
  24. KelvinW4

    KelvinW4 Registered Member

    Joined:
    Oct 11, 2011
    Posts:
    1,199
    Location:
    Los Angeles, California
    I'd doubt there'd be much I/O because it just checks for suspicious activity, process handling, but not scan and read everything you open and run.
     
  25. Brandonn2010

    Brandonn2010 Registered Member

    Joined:
    Jan 10, 2011
    Posts:
    1,849
    Ok. So are firewalls w/ HIPS and standalone HIPS programs generally light, or can they be heavy?
     
Loading...
Thread Status:
Not open for further replies.