Raynor

Member
  • Content Count

    99
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by Raynor


  1. Fair enough, but what's with fresh installations of v1803/v1809 ?

    According to MS, the memory integrity feature is always switched on
    on qualifying modern PCs (with virtualization support, UEFI and stuff)
    when Windows is installed from scratch.

    Wouldn't then "average" users be greeted by a big fat blue screen when they try to install EAM ?
    Or am I missing something here / am I getting something wrong ?


  2. Thanks for the confirmation and the quick reply.

    This should be documented somewhere to save others the hassle.
    E.g. in the release notes, as a sticky in the forum, or as a message
    in the installer.Ā  I was unable to find this info, which led to me being puzzled
    and wasting quite some time.

    Not a biggie at the moment, but compatibility with this featureĀ  would certainly
    be welcomed for the future. Other AV vendors (Kaspersky comes to mind)
    are also struggling with this feature, but they have been communicating
    it more openly.

    Thanks again
    Raynor


  3. There is a new security feature in Win 10 v1803 / v1089. It is called "Core Isolation". It can be found
    in the Windows Security Center under "Device Security". The core isolation feature includes a
    sub-feature called "Memory Integrity" (clicking on "core isolation details" reveals a switch that can be used
    to turn this feature on). It is enabled on fresh Windows installs, but not for existing installations that have been
    upgraded to v1803 or v1809. According to MS, these users can opt-in using the switch.

    For me, the switch turns on fine (no driverĀ  incompatibility warning given), but the required reboot ends with a
    blue screen
    KERNEL_SECURITY_CHECK_FAILURE. The welcome screen is shown forĀ  few seconds, then
    the BSOD is shown. I had to go into the BIOS, turn off virtualization, reboot and then disable the memory integrity
    setting in the registry.

    This happens on BOTH my PCs (main work PC - recent hardware, Z270 chipset - and my small Intel NUC7i5
    media PC with no special stuff installed). Tried it under Win 10 v1803 a couple of months ago
    and now again yesterday with v1809 (x64). Same results always.

    After pulling out some hair, I decided to uninstall EAM. And behold, the feature turns on successfully on BOTH PCs.

    Trying to re-install EAM with Memory Integrity turned on immediately causes the above mentioned BSOD
    during the installation
    (i.e. not on reboot, but immediately while the EAM installer is running).

    Here is another user reporting exactly the sameĀ  issue:

    https://www.wilderssecurity.com/threads/win-10-1803-core-isolation-and-memory-integrity.407342/#post-2776118

    "With Core Isolation and memory integrity turned on I got a green screen of death trying to install Emsisoft
    and could only recover using Macrium Reflect backup. Turned off memory integrity and EAM installed fine."

    Ā 

    The information given in the German section of the Emsisoft forum that it "should" be compatible
    is obviously FALSE. While EAM is certailny compatible with the basic "Core Isolation" feature,
    it does NOT work when the memory isolation sub-feature is switched on.

    https://support.emsisoft.com/topic/29479-windows-10-1803-kernisolierungspeicherintegritƤt/

    Botom line: please make it compatible šŸ˜

    Thanks and best regards
    Ranyor


  4. Dear all,

    I have one little improvement suggestion. Recently, I had do add quite a few program paths (mostly to .EXE files) to the exclusions in EEC
    (Exclude from monitoring) because the Behaviour Blocker behaved a bit overzealous on our client PCs.

    The problem with that was that most of the EXE files and paths that I wanted to exclude did NOT exist on our Windows Server 2016,
    as they pointed to programs that were only installed on (some) client PCs. But when adding an exclusion path in EEC, you are only
    given the chance to pick an existing file (in this case on the server).

    So I always need to use a workaround: First pick a file that exists on the server (e.g. "C:\Windows\notepad.exe" or whatever) and then
    manually change the path by clicking on it in the exclusions list and typing the real/desired path. This works as intended, the file is
    correctly excluded from scanning on the Client PCs. But all this is a bit cumbersome.

    So please let me kindly suggest that an option like "Manually add path" that allows to type in (or copy+paste from a textfile)
    any path (even to files that do not exist on the server)
    is added to EEC.

    Thanks and best regards
    Raynor

    Ā 


  5. We have bought 50 EAM keys for our corporate network.

    Not surprisingly, we have to decommission old PCs and replace them with new ones
    from time to time.

    Is it necessary to somehow remove the EAM license from the old PCs, or can
    the license key simply be reused on new PCs without us running out of activations
    (provided that, of course, the total number of PCs in operation at a given time does not
    exceed 50) ?

    Thanks and best regards
    Raynor


  6. On 7/26/2018 at 10:39 PM, Frank H said:

    The issue has been fixed in the upcoming EAMĀ 2018.7 release.

    Regarding this:

    I just have tested this in EAM 2018.6 stable) and i do not see those 2 options when i visit a site that sit in our blocklist. i.e. malwaretest.emsisoft.com

    I was logged in as a standard windows user and had 'read-only' permissions in EAM.

    Ā 

    Ā 

    Ā 

    Thanks for fixing!

    About the second issue:

    I saw these two options logged in as a local admin user with read-only permissions :)

    Ā 

    By the way: Is there any behaviour blocker test file/exe available, similar to the

    EICAR Ā AV test file ?


  7. OK, fair enough, I guess it comes down to personal preference. Everybody is entitled to their opinion,
    and I just wanted to express that I personally absolutely do not like the recent interface changes.

    Again, it's intended as constructive criticism, even if my wording might indeed have been a bit harsh (sorry about that šŸ˜‡).

    But on a factual level, I stand by my opinion šŸ™‚

    To give two concrete examples:

    1) The behaviour blocker program list in the protection section is in itself a scrolling list,
    so now we have a scrolling list view within a scrolling preferences list.

    2) The navigation tabs at the top (Behaviour Blocker, File Guard, Surf Protection) are still there,
    creating the illusion of a tabbed preferences window. But clicking on the tabs only scrolls down the
    view to the appropriate section. This just does not feel consistent/logical to me....

    • Upvote 1

  8. Well,

    the new scrolling settings view in 2018.06 already annoyed me (as well as many others), but back then I didn't feel the need
    to speak up because it was just one window. Now, with 2018.07 the scrolling view has been added to the "protection" section
    of the Interface as well.

    The usability of this scrolling view is HORRIBLE in my opinion, it is jumpy, makes you dizzy, and just somehow feels wrong.

    And as a bonus, I have added a little treat (see attached screenshot). On one of my PCs, I have increased the DPI scaling because I need bigger fonts.
    Now the main window is just a tiny little bit too large to fit the screen, which results in the rightmost scrollbar. But because of the scrolling settings,
    I now have TWO scrollbars. Now isn't that cute... no, wait, it isn't. The whole thing is just an abhorrent abomination.

    Bottom line: PleaseĀ  stop making the UI worse. There has been criticism by other users about the scrolling settings view in 2018.06 already.
    In fact, telling from the comments, nearly nobody seems to like it. Which is not surprising, because it's a bad design decision.

    The new On-Demand scanning default in 2018.07 is also a change to the worse, as Piotr has rightfully pointed out
    in his comments (especially in the second one) below the following article:

    https://blog.emsisoft.com/en/31683/new-in-2018-7-improved-file-guard-performance/

    Sorry for the scathing criticism (it is intended to be constructive!), but these recent changes indeed feel a bit like like making changes just for the sake of change... šŸ˜°

    Ā 

    All the best and best regards
    Raynor

    Two Scrollbars.png
    Download Image


  9. Thanks for your reply.

    I'm puzzled, because our whole domainĀ  setup is pretty much run-of-the-mill, nothing fancy...
    EEC uses its default ports, there are no fancy firewall settings in place, nothing.
    We use the update proxy (default port 8080) and that works fine as well.

    Some weeks ago, I installed EEC on the Server 2016, added the PCs, created manual deployment packages
    (we don't use remote deployment), installed them on the Win 10 v1709 clients via "Install.bat", and the connection
    to the Server worked flawlessly. On the clients no windows settings have been changed / no preparations have been made
    prior to deployment (if I understand correctly, preparations are only necessary for REMOTE deployment).

    Everything has worked fine for the last couple of weeks.

    Today, I upgraded some client PCs to v1803 through Windows Update (by removing the feature update deferral policy we had in place before),
    the upgrade itself went fine, but after that the clients all showed "connection" failed. Manually reconnecting was necessary and immediately
    worked fine (see above).

    I have just sent a mail to support asking for further instructions.

    I will be upgrading more clients tomorrow, so lots of logs will be there I guess :)


  10. System environment:

    Windows Server 2016 Domain
    About 40 domain-joined PCs, some Win 7, some Win 10
    EEC 2018.06 on the DC
    EAM 2018.05 (delayed feed) on the workstations

    Ā 

    First of all: installing and initially connecting EAM to EEC worked fone on all PCs.

    But:

    I have just upgraded a couple of our company's Win 10 v1709 PCs to Win 10 v1803
    using the normal Windows Update installation process.

    On all of the upgraded PCs, the connection to EEC is lost after the upgrade.

    This is annoying. After all the connection problems in the past (see this forum)
    I thought that these issues were a thing of the past. Obviously not.
    Connection stability sill seems iffy and unreliable, at least after upgrading
    domain-joined workstations from v1709 to v1803.

    I had to manually disconnect the PCs (using the interface within EAM itself, not
    from the server console) and then manually reconnected them (also from within EAM).

    After that, the connection was back up again.

    Still, unacceptable behaviour in a coroprate environment šŸ˜Ŗ

    Ā 

    Thanks and best regards
    Raynor

    Ā 


  11. On 3/18/2018 at 10:49 PM, Frank H said:

    Actually, since version 2018.2 there is a 3rd option in EAM and EEC: Autoresolve, notifications for threads only.
    This settings suppresses the Anti-Malware lookup notification and only shows a notification when a suspicious program was moved to Quarantine.

    https://blog.emsisoft.com/en/29745/new-in-2018-2-less-intrusive-smarter-notifications/

    However, settingĀ 'read-only' permissions for end-usersĀ  would be my advise anyways. An yep, EAM will run in auto-pilot mode for all alerts, also for theĀ  Behavior Blocker ones.

    I have now deployed EEC+EAM in our company.

    The read-only GUI setting works fine, but I really can't seem to find a way to make the behaviour blocker quarantine notifications read-only.

    Example: The behaviour blocker shows an alert because a suspicious program is trying to "change firewall settings".
    Users are told that the program will be quarantined after a couple of seconds, but are ALSO given the choice of clicking
    "Wait, I think this is safe"/"Thes program seems safe" (I can't remember the exact wording).

    I want to take that choice away from users, and NOT allow them to skip the behaviour blocker messages.
    They are the last line of defence, and I can guarantee that some users WILL allow new viruses to run by clicking "this is safe".
    Ā 

    So, please let me ask again: Is there any way to make the behaviour blocker prompts read-only without the option of skipping
    them ? I want them to work just as the normal "malware found" prompts (i.e. quarantining the program with no way around it).

    In there is no way, please let me kindly suggest again that an option to make these prompts read-only be added.

    Thanks and all the best
    Raynor


  12. On the topic of "playing around":

    I just installed a fresh copy of EEC (for pre-deployment testing purposes) in a fairly basic Windows Server 2016 virtual machine (VirtualBox),
    and stumbled over two obscure "Value does not fall within the expected range" warnings in the log.

    Again, quite basic & fresh Server 2016 VM, just installed EEC (2018.3.0.3338) a minute ago (first thing was taking a look at the logs).
    Any insights ?

    Best regards,
    Raynor

    Value.png
    Download Image


  13. 13 hours ago, GT500 said:

    VirtualBox has issues with security software (and always has), but that seems to have gotten worse lately. You can try the current beta version of Emsisoft Anti-Malware and see if it doesn't react as badly to some of the changes we've made, or you can add exclusions for VirtualBox so that Emsisoft Anti-Malware doesn't open hooks to its processes.

    Thanks, beta 2018.3.0.8528 seems to have fixed the problem.

    I had added an exclusion of the file VirtualBox.exe before, that helped as well.
    But with the recent beta, the exclusion is no longer required.


  14. Hi,

    starting yesterday (Friday 23-03-2018), VirtualBox always throws an error (see attached screenshot) when starting a virtual machine, be it an existing one
    or a new one. I was puzzled because I had extensively used VirtualBox just the day before (Thursday 22-03-2018) and everything worked fine (as it always had).

    This thread lead me onto the right track: https://forums.virtualbox.org/viewtopic.php?f=6&t=83791#p397261

    "The "process has more than one thread" error implies, I think, that some other process was trying to access VirtualBox process memory. So, either malware or antivirus. [...]
    It would be something invasive: system monitor of some kind, developer debug environment, crash manager maybe."

    Ā 

    And indeed, If I shut down EAM, VirtualBox starts to work again just fine (no errors whatsoever :rolleyes:).
    So some very recent protection/signature update must have made EAM more invasive, thus causing this error.

    Please fix this. :D

    I am using EAM 2018.1.1 (delayed feed), Win 10 x64 v1709, VirtualBox 5.2.8.

    Thanks,
    Raynor

    Ā 

    VirtualBoxError.jpg
    Download Image


  15. 1) Other context menus seem to be a bit wonky, too:
    Under "User Policies", If you right click on "Default for admins",
    you get a greyed out context menu, and if you right-click on "Default for non-admin",
    you get a normal, not greyed out context menu, but the "Edit" and "Delete" menu
    items do nothing. "Clone" works.

    Suggestion: Add a "New" Button at the bottom :), and make the
    behaviour for context menus on non-deleteable default groups consistent
    (i.e. consistently allow "clone" and grey out "edit" and "delete").
    Ā 

    2) Thanks, yes, that makes it clear.
    May I suggest that you rename this section from "Scanner Settings" to
    "Scanner settings for console-initiated scans" and/or add a little descriptive text like
    "These settings are applied when performing a manually initiated scan from EEC".
    This might save some confusion for future users.
    Ā 

    3) Thanks, this makes sense!

    Best regards,
    Raynor

    Ā