Jump to content
Welcome to InsanelyMac Forum

Register now to gain access to all of our features. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.


  • Content count

  • Joined

  • Last visited

About bengalih

  • Rank
    InsanelyMac Protégé
  1. Donk - thanks for those. Unfortunately, I've realized that my Dell PE 2900 III servers are not on the ESXi 6 HCL - so I might not be able to upgrade anyway. If anyone has tried ESXi 6.0 on PE 2900 server successfully let me know! But, I will ask for some more clarification: 1. Ok..I'm actually not using things like vMotion. I mostly use vSphere for templates/cloning and resource pools. I don't expect these things to work on OSX per-se, I just don't want to cripple this functionality for natively supported systems. As I mentioned an old issue with the unlocker seemed to lock/freeze my sessions to the server requiring me to restart some agents on the ESXi box. That seems to have gone away either in 5.5 or with the newer unlockers. I just wanted to make sure this stuff was clean. 2.Just curious about this and reading the article, along with my issue #4. If you turn smc.version = 0 doesn't that effectively disable the smc and not allow vCenter to manage? Or version is not the same as "present", I used the API path in the article you mentioned and the smc.present is showing as false on my 5.5 running unlocker. Are you saying that the newest version of the unlocker patches this so that vSphere sees an SMC and will allow me to start/stop systems from vSphere? If so...that's awesome (finally!) 3. Gotcha. I missed the "server" part. I don't need to run any versions of server before 10.7 so this won't affect me. 4. Ok that's great too. You might want to update the original instructions to note that so other people don't need to ask. Now I just need to decide I want to take the risk of upgrading to 6 without being on the HCL, or if i want to wait a while to see what others find!
  2. Hey Donk...thanks to you and Sam and everyone else getting this out. I've been holding off my upgrade to ESXi 6 until this was available. I do a lot of software testing on multiple versions of OSX and this makes it much easier. I'd like to ask some questions regarding this release and ESXi compatibility: 1) Is vSphere stable with this? I remember about 2 years ago there was an issue with vSphere stability that got fixed, just want to make sure I'm not going to have anything unexpected in this new one. 2) Could you give a brief description of what is/the necessity for "smcPresent vCenter compatibility" 3) Could you give a bit of insight on: "Added new EFI firmware files to remove Server check for 10.5 and 10.6". What is the need for this? I am pretty sure I built 10.5/10.6 images without any additional firmware on ESXi 5/5.5. Although it is *possible* that my original 10.5 was build with one of the firmware hacks and then cloned and upgraded to 10.6. I'm just trying to understand the need for these and why it only impacts 10.5/10.6 4) Regarding the VM HW level 11 issues. Is it that we can't initially create at 11, we can't upgrade to 11, or...? One fix is to change the HW level to 10, but that I believe is only possible on initial build correct? You can't downgrade from 11 to 10. Also, the other fix of smc.version = 0, does this mean we can run at 11 if we set this value? What (if any) negative impact does setting this to 0 have? thanks again - hopefully you have some time to answer so I can get a better understanding of the basics and can figure the rest out on my own
  3. Can someone explain this statement? I have a MacMini I would like to install ESXi 5.5 on and virtualize OSX 10.6 through 10.10. Can this be done successfully? If so, what are the drawbacks? From what I have read the main confusion was Fusion setting HW 11 to 10.10 boxes and ESXi 5.5 only recognizes HW 10. Specifically I'm not sure I understand what "once unlocked" means in this context? I would assume it means running the unlocker from this thread - but why is the unlocker needed to run on native hardware? I might just be reading the sentence wrong and it means either/or. Also, can someone else confirm that 10.10 will work on ESXi 5.5 as long as HW=10?
  4. I'm virtualizing a copy of OSX 10.6 through 10.9 on my ESXi 5.5 server using Donk's unlocker. I've installed the virtual tools that comes with that (Fusion tools i believe). Problem is I can't disable the Shared Folders option on the client. Since ESXi doesn't support shared folders, they aren't usable anyway, but it still creates a /volumes/VMware Shared Folders and you can see a shared folder quickly flash on the desktop at login. I've set the following .vmx option, that others have apparently had success with: isolation.tools.hgfs.disable= "TRUE" But, this doesn't work on any of the versions. Does anyone know how I can completely disable them on the OSX client?
  5. While I'm still awaiting the answer on the above, I have another (possibly related) issue? As I mention I'm running this on ESXi 5.5 and I continuously am getting the following errors in my system log: 2013-11-23 18:37:06 -0600 WindowServer[88]: _CGXGLDisplayContextForDisplayDevice: No matching context for device (0x7fcf60518790) - disabling OpenGL I've looked around and it seems like this is a common enough problem with no real solutions. I did however find this page: http://forums.macrumors.com/archive/index.php/t-1593194-p-4.html The poster there states this: "Ok, I have eliminated the _CGXGLDisplayContextForDisplayDevice error by removing the 10.7.2 GMA950 kexts, and entering 0x27A28086 into AppleIntelHD3000Graphics.kext/Contents/Info.plist, thus linking the 10.8.4 WindowServer & CGX library into a modern Mavericks native driver/kext." Now he is installing on actual older Mac hardware, not a VM...but I'm wondering if he is onto anything re: a solution. I'm still a bit of a neophyte when it comes to hardware on Macs, so just asking someone for input in translating this and if it might be applicable to VM as well to eliminate the error. TIA
  6. Hoping for some help...I was having some weird issues with my 10.9 install on ESX 5.5...it was going to the dreaded gray screen on logout and nothing could be done but reboot. Anyway.... I updated my ESXi 5.5 with all latest patches and updated the unlocker to 1.3. Then, I did a fresh install of 10.9 from the .dmg and everything looked fine. Installed VMWare Tools, rebooted...fine Installed guestd_patches.pkg, rebooted....fine Installed VMsvga2_v1.2.5_OS_10.9.pkg, rebooted...and now instead of a login screen I just have a black screen with a pointer. I restored a snapshot and I'm ok...but I'd like to have the VMsvga2_v1.2.5_OS_10.9.pkg installed if possible. Any ideas? UPDATE: I reapplied both patches to my reverted snapshot and the same thing occurs. Then, I tried to install just VMsvga2_v1.2.5_OS_10.9.pkg without guestd_patches.pkg. That resulted upon the grey apple screen on reboot with a stuck spinner...even after another cold boot. UPDATE2: Cold rebooted again from a stuck black screen and now it's stuck on the grey screen... so not sure if the guestd_patches.pkg had any real impact to the change in issues I'm seeing. UPDATE3: Safe mode boot (shift) works.. so likely a kext? Ok..no more updates until someone can offer me some advice. TIA! Ok...so I went ahead and applied the following to my .VMX file (from front page): svga.autodetect = "FALSE" svga.vramSize = "16384000" Now it doesn't really seem to say anywhere that applying these settings are mandatory, but rather an option. However I am sure that this is the difference between me having the issue and not. So...can someone confirm if these are supposed to be *required* for this to work and what exactly they are doing that would make the difference? Another update for posterity: I recently rebuilt all my OSX templates and didn't have any problems this time around. I think I built Mavericks with a different image this time (10.9.4 not 10.9.0), but I did not have to make any of the changes.
  7. This is possible...I believe I built my 10.6 box about 2-3 years ago from instructions gathered somewhere on this forum. I also believe that my 10.8 box was an upgrade from 10.6 and not a clean build. However, I really don't remember the procedure as it was so long ago. Could you tell me what/where I might check to see if this is the case? Also, if indeed I *did* build it that way - and it works without the unlocker, does that make sense? I mean if there is another way to get this going without the unlocker that could be of benefit especially to those of us who are trying to run with vsphere. thanks.
  8. So Donk/MSoK.... did I miss/misunderstand something here... I just rebuilt by 5.5 box. Did a total reinstall (not upgrade). Only thing that I kept was an array with a a VMFS datastore on it. After it was rebuilt I was able to boot my 10.6 and 10.8 boxes without installing the unlocker. 10.9 wouldn't boot - just stayed on the grey beachball screen. After installing the unlocker though, it booted. Is this known behavior that an unlocker isn't needed for 10.6/10.8 on 5.5??
  9. I read the information about only 4:3 being supported - and that's fine - but I guess what I'm wondering is why can't I go bigger than 1176x885? On my pre Mavericks I think I'm going something like 1400x1000 (can't remember exactly)...just know that it is bigger than I can get Mavs...
  10. Has the 10.9 patch been tested in ESXi? Specifically I'm running 5.5 and while the autofit patch seems to work, I don't seem to be getting any more sizes listed in display. The only two sizes listed when set to "Scaled" are: 1152x864 1024x768 If I maximize my vCenter window I can get it to read 1176x885 I am running this from the vSphere client on a system with a 1920x1080 resolution. Does this sound right?
  11. How are you able to downgrade HCV? I thought the only way to do this was to recover a snapshot or restore an older HCV from backup. Do you have a trick? Also, if you need any other testers on a pure ESXi 5.5 box, do let me know. Thanks again for all your work.
  12. Understood. I'm curious if you (or anyone else) is actually running ESX (or Workstation) on an actual Mac box? It would be interesting to know if you are required to choose HCV10 in order to choose 10.9 as the OS. That is the way it seems to be in Workstation 10 when I use your unlocker - just wondering if that is the way it would be even with Mac hardware? I would assume most people trying to use OSX as a production desktop will be running it on Workstation - so upgrading to HCV10 is not really an issue. For me, I like using the vSphere Client for ESX, so I don't want to move my HCV to 10. Being how 10.9 (and earlier) all seem to work fine for me - I'm ok. Of course, as I mentioned, I don't use OSX for a desktop - I use it for some software testing, and it seems to work well enough (I don't test things like usb, sound, ideal graphics, etc). Slightly OT - but is it possible that your update to the unlocker script that compresses the image could possible have fixed issues with vSphere connected ESX boxes? Since I reinstalled with the compressed image, things *seem* more stable to me.
  13. Ouch...9TB...wouldn't want to have to deal with that volume. Why not split it up and have each set of backup job go to a different Repository in Veeam? Of course, if you are backing up something that is 9TB for one server.... Yeah, personally I haven't had anything needing over HCV8. I didn't even realize I could go above that since I always use the vSphere GUI client. When I upgraded to 5.5 I realized I could go to 10, so I went ahead and just brought everything to 9 so I had flexibility in management. And yes...I know it would be better for me to mount my ISOs from another system over iSCSI or something, but this is just a SOHO lab setup and I don't have the resources to do that. I host my "home" DC off the ESX box which is my main file server (apart from my HTPC). So it shares out about .5 TB of data. I also have about a dozen lab systems setup. I have an old single core Athlon 64 3200 with 3GB of RAM that I run vCenter on (yes, I got it to run all components, including Update Manager on those specs!). I also have Veeam installed on that same box - including the SQL server to host both of them! It is slow to start, but it all works. I have about 2-3TB of disk on there which is enough to backup my ESX box. I might try to move my DMG to the Veeam NFS server so at least it is hosted on a different box and see if a build over GB Ethernet from a separate machine would also be problematic. However, as OSX is the only one that has ever had an issue, and now I've got my template built...not sure it is worth the hassle. Hopefully this info helps someone else out though...