|
You can run that off a NUC in a closet. I'm half-expecting that where it'll end up.
|
#
?
Dec 6, 2016 12:01
|
|
|
|
| # ? Apr 26, 2024 00:55 |
|
|
Is this just a limitation of HTML5 client in VCSA6.5 or do I need to do something special to drag VMs between/into folders? I get the (x) no matter what I do. HTML5 notes say "Migration" to folders doesn't work but I didn't think just moving a VM was migrating it. Works fine in Flash so I'm guessing I already know the answer but hoping someone can tell me otherwise.
|
|
#
?
Dec 6, 2016 12:55
|
|
|
If you're literally just doing syslog aggregation so you can grep through files later, whatever small specs are fine. If you're running something like Splunk, the Elastic stack, or Graylog so you can report on stuff, you can suck down plenty of memory on search indexes.
|
|
#
?
Dec 6, 2016 18:05
|
|
|
Computer Serf posted:Hmm bare bones built out is about half the price... meaning we could have a full cold backup for the same price as 24/7 enterprise support, so I guess this opens up an option for the suits to meditate on. Uhh, why not VMware Essentials? Do you have shared storage? "Cold backup" of a ESXi server is kind of a non-sequiteur unless you mean a DR site.
|
|
#
?
Dec 6, 2016 19:12
|
|
|
Has anyone tried enabling secureboot on 6.5? I seem to have a bunch of old vibs that I need to clean up because they aren't signed from 'esxcli software vib signature verify', mostly from old upgrade process, and NetApp needs to update their VAAI plugin.
|
|
#
?
Dec 6, 2016 20:35
|
|
|
I'm off on my first virtualization foray in my home / lab environment, and could use some goon advice with regards to data storage and backup. My setup is this :
What I want to accomplish is A. virtualize my poo poo. B. utilize my storage such that I have two sync'd file-level copies of all my data at any given moment, and C. have a versioned backup of it all. My stumbling block is where to set up the file-level type storage, versus where to set up the block level or system-image level backup. My plan is to install on the bare metal of Server 1 the MS Hyper-V Server 2016, and then virtualize my AD server instance on top of it. But I'm not sure where to store the VHD's - should I create an iSCSI target on my NAS? Or use one of my RAID partitions on server 1 for my VHDs? Also, because my AD server has (in the past) hosted things like redirected folders / user profiles, I'm not sure the best way to accomplish that going forward. Should I host redirected folders on an iSCSI NAS target mounted via my virtual AD server? Or use the bare metal hyper-v install to create a storage space on Server 1 that I allocate to my hyper-v VHD? And how does that play into how large my VHDs will be - and the best way to have both file-level redundant copies of my data on both my Server and NAS versus if I just had my virtual servers mount a shared SMB drive on the bare-metal Hyper-v server or something. Or maybe just abandon folder redirection altogether, because it is complete crap? Create an iSCSI target on my bare metal hyper-v host server that can be mounted by my virtual servers, rather than setting that up on my NAS? I think all the versioning can be handled by Synology backing up a hyper-backup image to my external HDD - but also open to suggestions on this... mindphlux fucked around with this message at 09:13 on Dec 7, 2016 |
|
#
?
Dec 7, 2016 06:42
|
|
|
We are running third party Samsung 850s in our fleet of vdi hosts as local cache. The hosts are dl380g9s. Make sure you install the HP driver for the sas controller.
|
|
#
?
Dec 7, 2016 21:22
|
|
|
Has anyone tried to do something stupid like use vSphere Replication manager to replicate VM's to a separate datastore within a datacenter that are then synced offsite either via replication or scheduled job to AWS S3? I'm trying to fulfill a requirement to get VM data out of our datacenter as agnostically as possible. So far I've configured a replication using VRM to replicate VM's to a datastore that is backed by Amazon Storage Gateway, which copies them up to AWS S3. I'm struggling on the restore part of it, as they want to restore the VM's into EC2. As far as I can see VRM uses a fairly proprietary format for the destination replica VMs that most likely won't work with the ec2 cli import feature, as that requires the VM to have been exported as an OVA files instead. I figure this is probably why people charge and arm and a leg for such replication technology.
|
|
#
?
Dec 7, 2016 23:53
|
|
|
The only way I can think of doing this would be to restore to EC2 via your vCenter server using something like https://aws.amazon.com/ec2/vcenter-portal/ If you want to keep a replica of your VMs in a cloud provider ready to boot up for DR then your best bet is to find a service provider that can offer you Veeam (or build it out yourself) https://www.veeam.com/cloud-connect.html
|
|
#
?
Dec 8, 2016 00:03
|
|
|
Funnily enough we've looked at all those options and had them shot down by upper management  Hopefully management will stop smashing themselves in the foot soon.
|
|
#
?
Dec 8, 2016 04:32
|
|
|
About six weeks ago, Amazon/VMware announced VMware Cloud on AWS. If you have an AWS rep already, it's worth seeing if they can get you early access.
|
|
#
?
Dec 8, 2016 05:03
|
|
|
gently caress Xenserver. That's all.
|
|
#
?
Dec 8, 2016 18:36
|
|
|
hihifellow posted:gently caress Xenserver. Sounds about right.
|
|
#
?
Dec 8, 2016 19:29
|
|
|
hihifellow posted:gently caress Xenserver. 
|
|
#
?
Dec 9, 2016 00:26
|
|
|
Vulture Culture posted:About six weeks ago, Amazon/VMware announced VMware Cloud on AWS. If you have an AWS rep already, it's worth seeing if they can get you early access. Yeah, I was joking with my boss that AWS and VMware timing is absolutely horrible since we had this DR project foisted on us by an audit and if it could be pushed back by 6 months it would probably eliminate the need for 2-3 months of work
|
|
#
?
Dec 9, 2016 03:19
|
|
|
evol262 posted:Uhh, why not VMware Essentials? Do you have shared storage? "Cold backup" of a ESXi server is kind of a non-sequiteur unless you mean a DR site. We have some Synology units, but they're basically already allocated for specific uses. It does seem pretty silly to not set up HA, granted we get a cloned set of hardware. adorai posted:We are running third party Samsung 850s in our fleet of vdi hosts as local cache. The hosts are dl380g9s. Make sure you install the HP driver for the sas controller. Nice to know someone out there is trail blazing. We're just a smallish office with our NOC on the same floor so I think just going all in on a fully supported branded Enterprise™ setup may be the wisest choice. The HA setup with some DIY elbow grease is really tempting but having never dealt with VMware before if poo poo hits the fan I don't want to be the one solely responsible for setting up the fan and causing poo poo to hit it.
|
|
#
?
Dec 9, 2016 03:57
|
|
|
We would have gone enterprise if it wasn't just cache. Losing the cache on a single host would be a non event.
|
|
#
?
Dec 10, 2016 17:35
|
|
|
Anyone have much experience with VSAN in production environments? I'm wondering what kind of issues I can expect to run into with it, any gotchas etc?
|
|
#
?
Dec 13, 2016 01:29
|
|
|
theperminator posted:Anyone have much experience with VSAN in production environments? Make absolute sure any raid/passthrough cards you have are explicitly said to be compatible with VSAN by VMware.
|
|
#
?
Dec 13, 2016 02:37
|
|
|
Methanar posted:Make absolute sure any raid/passthrough cards you have are explicitly said to be compatible with VSAN by VMware. Seconding this. The HCL is critical to vSAN success. http://www.vmware.com/resources/compatibility/search.php?deviceCategory=vsan
|
|
#
?
Dec 13, 2016 14:31
|
|
|
How do I get rid of orphaned vms? In my home lab (esxi/vcenter 6.5) I tried deploying an ova (sexilog) and it failed. The vm shows up as orphaned in my inventory but I can't find any options for deletion.
|
|
#
?
Dec 14, 2016 00:44
|
|
|
Cool, I've checked that our R630s with PERC H730P's are supported, we will have 10G ports dedicated to VSAN traffic & our servers have tonnes of RAM & CPU so we should be all good from that point of view.GobiasIndustries posted:How do I get rid of orphaned vms? In my home lab (esxi/vcenter 6.5) I tried deploying an ova (sexilog) and it failed. The vm shows up as orphaned in my inventory but I can't find any options for deletion. What I usually do in this case is create a new folder, move the VM to that folder then delete the folder.
|
|
#
?
Dec 14, 2016 00:50
|
|
|
theperminator posted:What I usually do in this case is create a new folder, move the VM to that folder then delete the folder. This worked, thank you! Now when I try to deploy I get the message 'ovf package with compressed disks is currently not supported for ovf import' so that's a whole new problem, but at least I don't have orphaned vms sitting in the inventory.
|
|
#
?
Dec 14, 2016 01:29
|
|
|
If I want to move a template between vCenters, what's the best way to do that? If it's as easy as copying the folder from one datastore to another, how do I get vCenter to see the new template?
|
|
#
?
Dec 14, 2016 02:02
|
|
|
anthonypants posted:If I want to move a template between vCenters, what's the best way to do that? If it's as easy as copying the folder from one datastore to another, how do I get vCenter to see the new template? I just browse the datastore to the template's location, right click on the.. uh.. I think it's vmtx file and it should give you the option to register it with vCenter. I should remember better seeing as how I did this not more than two weeks ago.
|
|
#
?
Dec 14, 2016 02:32
|
|
|
theperminator posted:Cool, I've checked that our R630s with PERC H730P's are supported, we will have 10G ports dedicated to VSAN traffic & our servers have tonnes of RAM & CPU so we should be all good from that point of view. We've had some issues with VSAN deployments. Issues with supposedly supported pci flash on Cisco servers not working. HCL approved hardware throwing random, unreproducible errors. On one deployment it randomly decided that SSD drives were actually HDD drives. Hosts that won't go into maintenance mode. Seeming false negative health alarms. I also hate that it breaks the fundamental "VMs are files" construct that makes VMware so flexible. You can't just jump on a host and copy things around if they life in a VSAN datastore because they aren't really files anymore, they are objects pretending to be files. It can make recovery painful in certain situations.
|
|
#
?
Dec 14, 2016 22:08
|
|
|
I'm about ready to get a gas can and light these X710's on fire, the firmware and drivers are absolute garbage. Hopefully Dell will bite on swapping them out for X520's, I've stumped the majority senior support people with getting these stupid things to handle LACP correctly.
|
|
#
?
Dec 14, 2016 23:50
|
|
|
BangersInMyKnickers posted:I'm about ready to get a gas can and light these X710's on fire, the firmware and drivers are absolute garbage. Hopefully Dell will bite on swapping them out for X520's, I've stumped the majority senior support people with getting these stupid things to handle LACP correctly. That's where we were at and we basically told our rep to give us X520s or they were getting the whole order back.
|
|
#
?
Dec 14, 2016 23:52
|
|
|
Maneki Neko posted:That's where we were at and we basically told our rep to give us X520s or they were getting the whole order back. Same, but we've concluded that PXE booting in UEFI mode with X710s is basically impossible.
|
|
#
?
Dec 15, 2016 01:45
|
|
|
Is anyone else seeing VMs misreporting virtual memory utilization after the 6.5 upgrade? I have a couple that are wildly incorrect beginning the second they migrated to a 6.5 host. That particular VM only has 2gb out of 5 even allocated inside the OS memory manager, there is no way the active pages are that high. I have at least a half dozen doing this and they all show that steady increase since I did the upgrade on the 10th.
|
|
#
?
Dec 15, 2016 16:39
|
|
|
I must be missing something with virt-install, but after failing to boot an Ubuntu cloud image I wanted to try to create one from scratch following http://docs.openstack.org/image-guide/ubuntu-image.html to see if it that would work better, but I get the exact same behavior: instead of booting it's just stuck at the "bios" screen. Anyone know what the problem could be? I'm using a 32 bit 16.10 image inside another VM (CentOS) if that might have anything to do with it. Not sure what more info to provide, I'm running the command as in the openstack docs when creating the VM with the ISO and with "--import" when trying to use the cloud image.
|
|
#
?
Dec 15, 2016 17:35
|
|
|
netcat posted:I must be missing something with virt-install, but after failing to boot an Ubuntu cloud image I wanted to try to create one from scratch following http://docs.openstack.org/image-guide/ubuntu-image.html to see if it that would work better, but I get the exact same behavior: instead of booting it's just stuck at the "bios" screen. Anyone know what the problem could be? I'm using a 32 bit 16.10 image inside another VM (CentOS) if that might have anything to do with it. Is tested virt enabled?
|
|
#
?
Dec 15, 2016 18:01
|
|
|
evol262 posted:Is tested virt enabled? No idea so I'm guessing not  How/where do I check? I just went home so I'll have to do that tomorrow though.
|
|
#
?
Dec 15, 2016 18:16
|
|
|
Should have been nested. Autocorrect. You can check in /proc (Google will tell you exactly where to look, depending on whether amd or Intel)
|
|
#
?
Dec 15, 2016 18:37
|
|
|
Welp, finally starting to run into some of these vaunted VVOL "bugs". Have Nimble arrays running the 3.4 firmware at two sites (Production & HQ) one side running mostly vSphere 6.0 U2 and the other running vSphere 5.5/5.1 and some 6.0 U2. When trying to vMotion a VVOL stored virtual machine in our production side, I always get an error "sms.fault.VvolRebindFault", even when the VM is powered off first. The only workaround we have right now is to Storage vMotion to a non-VVOL datastore first, which lets us then use VMotion normally. VMware support has been less than helpful, saying that they are aware of this bug but the only "fix" they can offer is to update the VASA storage provider to the latest version. Nimble support is even more confused at this point than VMware's support is.
|
|
#
?
Dec 15, 2016 21:43
|
|
|
Wicaeed posted:Welp, finally starting to run into some of these vaunted VVOL "bugs". Heh, I have not bothered to migrate to VVOLs yet on our Nimble SANs yet. Guess that is a good thing!
|
|
#
?
Dec 15, 2016 21:55
|
|
|
Moey posted:Heh, I have not bothered to migrate to VVOLs yet on our Nimble SANs yet. Guess that is a good thing! More troubling is that it's only an issue on our Prod side, our Office HQ side is fine  I've been lukewarm with VVOL's so far. When they were working, it was fine, but there were some small issues that got at me, such as when a VM's vmdk volume space usage was beyond 80% it will generate Nimble alerts of the same severity as a non-VVOL datastore being over it's quota or reserve, which is irritating. More irritating is that Nimble considered this a non-issue, despite the fact that you can't individually set the quota alert threshold to 100% without manually editing each single disk from the Nimble UI. Or the fact that due to space calculated you will often see a vmdk usage of OVER 100%, which generates nasty alarms. Once again Nimble considered this a non-bug 
|
|
#
?
Dec 15, 2016 22:40
|
|
|
Martytoof posted:I just browse the datastore to the template's location, right click on the.. uh.. I think it's vmtx file and it should give you the option to register it with vCenter.
|
|
#
?
Dec 15, 2016 22:51
|
|
|
BangersInMyKnickers posted:Is anyone else seeing VMs misreporting virtual memory utilization after the 6.5 upgrade? I have a couple that are wildly incorrect beginning the second they migrated to a 6.5 host. From my understanding the vmware memory graphs won't generally match what the OS reports because the hypervisor is unaware of freed memory by the guest. We've found for instance that our Java Applications only use 2GB out of the available 8, but after running for long enough with enough Garbage Collection happening, all of the memory gets touched and the memory usage of the guest is reported as 100% by vmware What does the "Active" metric look like for your VMs?
|
|
#
?
Dec 15, 2016 23:59
|
|
|
|
| # ? Apr 26, 2024 00:55 |
|
|
At home last weekend i tried to install linux mint in an oracle virtualbox vm, as a place for loving around with programming stuff (trying out .net core and vs code at the time) without dualbooting out of Windows. Gotta play my games and all. Problem was the linux mint desktop kept turning into a glitchy godawful mess if i had 3d accel on, and dogshit slow with it off, and I didn't have a lot of time to mess around right then so I haven't got back to it. Anyway, question is, what do people here like as far as free Windows-based virtualization hosts, and guest linux OSes for said? Is Oracle Virtualbox the only free game in town? (I know there's ESXi on the bare metal side.)
|
|
#
?
Dec 16, 2016 00:08
|
|
