|
I'm definitely a supporter for moving towards using Workstation (and Fusion?) to manage vSphere. I have Workstation on every system I use on a regular basis and often have my own VMs that I need to run for desktop productivity, anyway. Consoles tend to open with a pretty noticeable delay for me, which is a bit frustrating. I avoid it not with RDP, but a bunch of remote management console tools or well, remote terminal/SSH for the Linux machines. It's probably a more elegant approach anyway, which is why RDP/VNC/Console access was never really pressed hard by consumers and customers alike. I feel a faster console would be nice, either way. And perhaps a nod to those who can't set up for much in the way of remote/centralized Guest OS-specific management. Honestly, a lot of my own interests/hopes at this stage are closer to a niche segment of gamers who would like to virtualize. I would prefer to pour my resources into the home lab and stop building gaming PCs altogether - just equip a server very well with lots of RAM and a good video card, then set up a good VM as the gaming desktop, and rely upon a thin client. There's sufficient separation, resource sharing and priorities, and some of the physical and logical configuration requirements are greatly simplified, I can argue. Most would just KISS and stick with a separate gaming system, so I'm definitely hopeful of something that most would really not want to deal with.  Back to console bits, I had an Nvidia GPU running very respectable frame rates for a GL benchmark app/demo on ESXi but could not actually see/enjoy the benefits so well, due to draw rates for the remote console on my desktop. Local LAN, 1Gbit. And I also have a real desire for continued improvements and even a new approach to Workstation/Fusion/Type-2 hypervisor GPU support and virtualization - allow Guest OSes to actually see an instance of the installed video card (with appropriate abstraction and separation). Is that crazy at all? I wonder if using an "unsupported" virtual adapter (as we do now) within VMs is really enough. For instance, you will have zero luck getting MacOS running perfectly in a VM due to Quartz not being supported with the virtual adapter.
|
#
?
Feb 13, 2014 07:46
|
|
|
|
| # ? May 10, 2024 06:35 |
|
|
mayodreams posted:I was checking that course out, and I can't tell if it online or local only, and/or how much it is. Entirely online. $185
|
|
#
?
Feb 13, 2014 10:00
|
|
|
Have any of you used Double-Take Availability for vSphere? We use their availability software withing our server instances for some file servers and a DB server and it works pretty well. Just wondering if anyone had a good or bad experience with their vSphere plugin.
|
|
#
?
Feb 13, 2014 15:47
|
|
|
Kachunkachunk posted:I'm definitely a supporter for moving towards using Workstation (and Fusion?) to manage vSphere. WS/Fusion will never be able to fully drive vSphere, but we are getting more and more with each release. If you have specific feature request, give them to the PM. If you don't know who he is, email me. Which console version were you using? Connecting from WS to ESX, I can play all sorts of games. Skyrim even works as long as the setting are down and you don't get too many characters on screen at once. I was playing though some gunpoint levels the other day at work. It was "testing" the product. Yes, it is crazy to want to see an instance of the physical graphics card. The appropriate abstraction is the VMware SVGA device. We will already advertise different capabilities in the guest based on what the hardware can support. We are able to keep cool stuff like vMotion working with that because we have different capability levels. When you start doing the shared passthough, you lose most of your VM features. Stuff as simple as suspend/resume go away. We have been loosely supported as a legit graphics card in Windows for a long time. We were getting certified as an unclassified driver. Microsoft would only allow virtual graphics drivers as unclassified. Now, they won't allow that. That means that we have to have a driver that meets the DirectX spec, a device that handles it correctly, and be able to render all of it correctly. That is every bit as supported as any physical card. As far as MacOS, the virtual device is fine. The driver is what needs work. That simply hasn't been done because there is no business reason. The 2D graphics is good enough for now. We actually had a guy put a few months of work into it last year, but full 3D support would take several months more. We just don't have the time for that when we have tons of people needing a Windows driver that is certified and can run all the CAD apps they have.
|
|
#
?
Feb 13, 2014 16:43
|
|
|
Some sales guy from VMTurbo managed to get me. I see 408 numbers and answer them out of habit. He talked me into the dog and pony demo show. Anyone know anything about their product?
|
|
#
?
Feb 13, 2014 23:30
|
|
|
I've got an issue with 4TB disks on my ghetto ESXi 5.1 whitebox.. I have a local SATA disk, and have set up the drive as an RDM (vmkfstools -z), but if I add the disk to my Server 2k8r2 VM, the partition table corrupts and I lose the disk, it only seems to see 512B. I've tried partedUtil -fixGpt but tells me there is no GPT partition. It seems this is a fairly common problem, just wondering if anyone has a simple solution. I've tried adding the disk as a datastore (VMFS-5) instead but it still won't let me create a 4TB disk - max 2TB. My fallback is to just create two 2TB disks and move on with my life but I would prefer to have it as an RDM (pure NTFS) disk if at all possible. Is there something obvious I am missing?
|
|
#
?
Feb 14, 2014 12:04
|
|
|
skipdogg posted:Some sales guy from VMTurbo managed to get me. I see 408 numbers and answer them out of habit. He talked me into the dog and pony demo show.
|
|
#
?
Feb 14, 2014 13:41
|
|
|
kyojin posted:I've got an issue with 4TB disks on my ghetto ESXi 5.1 whitebox.. I thought datastores larger than 2TB was a feature of 5.5
|
|
#
?
Feb 14, 2014 17:59
|
|
|
thebigcow posted:I thought datastores larger than 2TB was a feature of 5.5 Datastores have been able to be much larger than that for quite some time, but individual extents prior to VMFS5 could not be larger than 2TB.
|
|
#
?
Feb 14, 2014 19:29
|
|
|
thebigcow posted:I thought datastores larger than 2TB was a feature of 5.5 I believe that GPT & 2TB+ support came in VMFS-5 with ESXi 5.0? I would try 5.5 but it won't work with my hardware. The drive is correctly listed under ESXi Config>Storage>Devices as 3.64TB though which suggests that vmware is correctly recognising it. It is plugged directly into the motherboard - Gigabyte GA-990FXA-UD5. The disk is a Seagate 4TB, not sure of the specific model but I can find out if need be. Here are the results of my playing about with it all day.. I created a ubuntu machine and installed Gparted - when adding the drive initially, it complains that the backup GPT partition is corrupt but the primary is ok - if I apply this, Gparted will crash. If I cancel it and create a new partition that works with no problems. I can then create a 4TB NTFS volume in Gparted and if I mount this it seems to be fine - I can create folders and documents (only small files, nowhere near the 2TB boundary), and it survives a reboot of either VM or the entire host. Doesn't automatically mount following a reboot but no doubt I could fix that with fstab if needed. If I remove the disk from the Ubuntu VM and add it to the Server 2k8r2 VM, I can see the volume and the folders created and I can create new files and folders. Disk properties>Volumes tabs reports the capacity as 0MB with unallocated space of -3815446MB which doesn't seem ideal.. however Volume List (at the top of the Disk Management window) and Windows Explorer both report the capacity correctly if slightly differently, 3,726MB in Disk Management and 3.63TB in Explorer. I can offline the disk and bring it back online and everything is still there. However, when I remove the disk from the Windows VM and add it back to the Ubuntu VM, the volume is lost and GParted again reports the partition table as corrupt. If I reboot the Windows VM with the disk attached, when windows comes back up it tries and fails to initialize the disk (error says the Disk Management View is not up to date) and the volume is gone. Seems like it could be a Windows issue, or something to do with the PVSCSI driver? I have VMware Tools installed on that machine, although I also tried without and got the same result. Tomorrow's plan is to scrap ESXi and install windows directly to the hardware I think, see what I get then. Possibly a hardware/controller issue, but I upgraded to the 990FXA specifically to deal with that so I hope not. Any suggestions would be really helpful - I am by no means a partitioning expert and this seems completely bizarre to me.
|
|
#
?
Feb 14, 2014 20:09
|
|
|
Workstation question for a change. If I want to try to put two instances of Workstation, on 2 separate computers, on the same subnet, what's my approach? One has a host only network of 10.1.10.0/24 on it, as does the other, and I'd like them to connect in such a way that they're both on the same 10.1.10.0/24. This is probably suicide, no?
|
|
#
?
Feb 14, 2014 20:37
|
|
|
Depends on why you're doing it. If you want to keep the host-only networks, then you would have to configure the two hosts as routers and set up a VPN tunnel between the two guests, which might be useful experience. If you just want the VMs to be able to communicate with as little work as possible, then set the network configuration to Bridged and give the guests the 10.1.10.0/24 addresses you want and that's it. You'll still need to manage routing if the hosts are on a different subnet and you want the guests to be able to reach them though. e: As an aside, bear in mind that an IP address and "being on the same subnet" are layer 3 issues, but it sounds like your question is actually about layer 2 networking. A host-only network is a virtual layer 2 network independent from the one the host is physically connected to, which is why you need routing at layer 3 to join them together. A bridged virtual network makes the host act as a switch and connects the virtual layer 2 network to the physical one. So if you do this and give your VMs 10.1.10.0/24 addresses, while they will be on a separate 3 network and (assuming your hosts are on a different subnet) can't communicate with the hosts, this is only because you've addressed them that way. They will still see broadcast traffic from the rest of the network and you would have to add them to a separate VLAN if you want them to really be on their own network. Scikar fucked around with this message at 21:12 on Feb 14, 2014 |
|
#
?
Feb 14, 2014 20:58
|
|
|
We've got a proper DEV environment, but for just sketching ideas, some of our developers have a few VMs set up on their box. I feel like I don't want to change them to bridged because I don't really want these VMs on the network proper, so of the two options, it seems like setting up a VPN might work in this case. I just want to try to tie two of the developer PCs together. And as you said, useful experience.
MC Fruit Stripe fucked around with this message at 21:14 on Feb 14, 2014 |
|
#
?
Feb 14, 2014 21:12
|
|
|
MC Fruit Stripe posted:We've got a proper DEV environment, but for just sketching ideas, some of our developers have a few VMs set up on their box. I feel like I don't want to change them to bridged because I don't really want these VMs on the network proper, so of the two options, it seems like setting up a VPN might work in this case. I just want to try to tie two of the developer PCs together. And as you said, useful experience. ipsec tunnel is right
|
|
#
?
Feb 14, 2014 21:19
|
|
|
Look up best practices for 10gig networking. It goes over all of thsiMisogynist posted:Oh God, these fucks kept calling the CIO at my last job four times a week before they finally got a hold of my number and I set their behavior straight. I have never met a company with more aggressive direct marketing. They carpet bombed my entire organization (bothering at least 20 people that I know of) for weeks. They also went after our C levels. Raising the issue with them only got them to slow down. The spam/cold calls continue to this day just at a reduced rate. I couldn't get past the awful sales weasels to actually give it a fair shake.
|
|
#
?
Feb 14, 2014 21:23
|
|
|
parid posted:I have never met a company with more aggressive direct marketing. They carpet bombed my entire organization (bothering at least 20 people that I know of) for weeks. They also went after our C levels. Raising the issue with them only got them to slow down. The spam/cold calls continue to this day just at a reduced rate.
|
|
#
?
Feb 14, 2014 21:31
|
|
|
Misogynist posted:I once literally made a cold caller cry so You are doing God's work 
|
|
#
?
Feb 14, 2014 21:33
|
|
|
My go to reply for cold callers is that we have no budget for IT expenditures the next three years and they usually just say "oh ok bye" and never call back. If they push me I ask them if they're volunteering to do free work for us and they hang up.
|
|
#
?
Feb 14, 2014 22:06
|
|
|
Welp congrats to the new VCDX's! Glad to see my friend got his!
|
|
#
?
Feb 14, 2014 23:13
|
|
|
Dilbert As gently caress posted:Welp congrats to the new VCDX's! They announced results? I thought the grading rubric usually takes a week.
|
|
#
?
Feb 15, 2014 00:05
|
|
|
El_Matarife posted:They announced results? I thought the grading rubric usually takes a week. https://twitter.com/herseyc  He told me he's 128
|
|
#
?
Feb 15, 2014 00:20
|
|
|
I'm having a hard time finding it, but wasn't there a thing a little while back about default passwords expiring on SSO or something? I'm going through the VCP-DCV class now so a lot of the gobbledygook for earlier is starting to click and I'm thinking "poo poo, I didn't understand at time but it sounded important!" Hopefully this sounds familiar to someone, my memory is bad because my comprehension at the time was bad.
|
|
#
?
Feb 15, 2014 03:54
|
|
|
Dr. Arbitrary posted:I'm having a hard time finding it, but wasn't there a thing a little while back about default passwords expiring on SSO or something? The VCSA has a hard 90 day root password issue that can lock you out. SSO is hard locked to the original password if you want to fail back or recover. If you need help studying let me know I don't mind doing a talk on the VCP test. Really do not mind it.
|
|
#
?
Feb 15, 2014 04:01
|
|
|
Dilbert As gently caress posted:If you need help studying let me know I don't mind doing a talk on the VCP test. Really do not mind it. I plan on taking the 5.1 exam and I appreciate all the help I can get. I've been reading Hersey's blog and the Scott Lowe book. I've got the blueprint printed out and I'm trying to figure out what areas I need help with. I got a 66% on the practice test, I guess that's not bad for now. Are there any really good practice exams that you can recommend?
|
|
#
?
Feb 16, 2014 02:23
|
|
|
El_Matarife posted:They announced results? I thought the grading rubric usually takes a week. You can find out the same day. I got my results the next day when I did it.
|
|
#
?
Feb 17, 2014 02:04
|
|
|
Who the hell thought it was a good idea to not add support for virtual hardware version 10 to the Fat client?
|
|
#
?
Feb 17, 2014 13:43
|
|
|
Mausi posted:Who the hell thought it was a good idea to not add support for virtual hardware version 10 to the Fat client? It was intentional, part of their way of phasing-out the fat client (All new features in 5.5 and later are not available in the fat client).
|
|
#
?
Feb 17, 2014 14:23
|
|
|
Not implementing advanced new functions I can understand, but being able to edit a VM config is just about the most basic function the client could have. loving ridiculous decision, especially when you consider that you still need to use the fat client for loads of things like SRM, so you can't easily work on both concurrently.
|
|
#
?
Feb 17, 2014 14:37
|
|
|
Yeah I agree it is ridiculous. DevNull, who works for VMware, has spoken about it from the remote-console side of things earlier in the thread. I haven't used the web client much so far but from my brief poking around it really doesn't feel like something I'd enjoy using on a day-to-day basis (Especially as our vCenter server is garbage). Is there VUM support in the web client? Because if not that's another annoyance.
|
|
#
?
Feb 17, 2014 14:42
|
|
|
Mausi posted:Who the hell thought it was a good idea to not add support for virtual hardware version 10 to the Fat client? Yeah, this is ridiculous. Is there even a way now to manage all the features of a free ESXi install?
|
|
#
?
Feb 17, 2014 14:45
|
|
|
HalloKitty posted:Yeah, this is ridiculous. Is there even a way now to manage all the features of a free ESXi install? Nope (AFAIK). Apparently you can use Workstation to do some things but other than that you need vCenter (Dilbert As gently caress: feel free to correct me).
|
|
#
?
Feb 17, 2014 15:07
|
|
|
cheese-cube posted:Is there VUM support in the web client? Because if not that's another annoyance.
|
|
#
?
Feb 17, 2014 15:11
|
|
|
Mausi posted:Nope! Fat client only. I can't wait to see what happens when I attempt to use plugins like VUM or NetApp VSC against the new VMs. We run NetApp VSC in our environment so please report-back when you have a chance (We are only running 5.1 at the moment and an upgrade to 5.5 is a long way off).
|
|
#
?
Feb 17, 2014 15:13
|
|
|
cheese-cube posted:We run NetApp VSC in our environment so please report-back when you have a chance (We are only running 5.1 at the moment and an upgrade to 5.5 is a long way off). Well I can tell you a couple of things, because I've already completed the integration with the Fat Client/vCenter 5.5 1.) Automated install is a PITA, because you have to complete a pop up webpage to register it with vCenter before the roles are added. What we've done is pause the install script to wait for the user to complete the registration before continuing on to do things like automated user priviledge and role management. 2.) It is working with 5.5 hosts, reporting on best practice settings and all that. I haven't yet tried scheduling snapshots with it, but that's all we use it for. I wouldn't like to think what would happen if you tried to do automated backups or disk alignment with it.
|
|
#
?
Feb 17, 2014 15:29
|
|
|
Mausi posted:Well I can tell you a couple of things, because I've already completed the integration with the Fat Client/vCenter 5.5 Thanks for that info, I'll keep it in mind should a 5.5 upgrade appear on the road map. Regarding automatic alignment, I would strongly recommend against enabling it in your environment unless there are noticeable issues with NFS I/O latency. I may have posted this earlier in the thread or over in the Enterprise Storage Megathread but the automatic alignment feature was causing havoc in our environment due to snapshot-removal stun times on VMs.
|
|
#
?
Feb 17, 2014 15:35
|
|
|
cheese-cube posted:Regarding automatic alignment, I would strongly recommend against enabling it in your environment unless there are noticeable issues with NFS I/O latency. I may have posted this earlier in the thread or over in the Enterprise Storage Megathread but the automatic alignment feature was causing havoc in our environment due to snapshot-removal stun times on VMs. Good tip. We don't use it at all - if anything we're too far the other way where we only check on disk alignment if a volume or array starts showing performance issues. All recent OS builds are aligned on creation and you can't create a VM outside of the automated build system (or I
|
|
#
?
Feb 17, 2014 15:39
|
|
|
So am I misunderstanding, or does installing vCenter in a greenfield deployment now literally involve creating a version 9 VM and then upgrading it later?
|
|
#
?
Feb 17, 2014 18:25
|
|
|
Misogynist posted:So am I misunderstanding, or does installing vCenter in a greenfield deployment now literally involve creating a version 9 VM and then upgrading it later? I suppose in theory you could create a ver10 VM using Powercli, but I wouldn't be happy with it.
|
|
#
?
Feb 17, 2014 18:32
|
|
|
cheese-cube posted:We run NetApp VSC in our environment so please report-back when you have a chance (We are only running 5.1 at the moment and an upgrade to 5.5 is a long way off). You can play around with the VSC web in VMWare Hands on Lab HOL-PRT-1301. It very different as the functionality is largely integrated into the normal views, rather than being segregated into NetApp specific pages, so it can take a big of time to figure out things like modifying backup schedules, or listing available snapshots for restore. It should, however, be faster than the older VSC plug-in, which was painfully, abysmally slow if you were backing up a large number of VMs and datastores. VSC 5 is still only an incremental improvement though. The big changes are coming with VSC 6.
|
|
#
?
Feb 17, 2014 21:23
|
|
|
|
| # ? May 10, 2024 06:35 |
|
|
cheese-cube posted:Yeah I agree it is ridiculous. DevNull, who works for VMware, has spoken about it from the remote-console side of things earlier in the thread. The remote console is the same for both the thick client and the web client. We made a bunch of improvements to the it, but sadly they don't get seen much. Mostly because the web client sucks so much. I think part of the problem is also that the remote console was so bad for so long, that people are afraid to even touch it. It was neglected for years. The only up side is that we were able to implement a remote console that is HTML5 based. That adds a lot of flexibility, so you can just jump on any browser and get a console without an install. We will be keeping the installed console around though. It is better performance, and will always have better support for mouse and keyboard interaction.
|
|
#
?
Feb 17, 2014 22:43
|
|