VMware vSphere 6.5 released!
It's here! Finally, those GA bits arrived today, ready for you to download. My original article has now been updated to include the new download links:

When doing both files concurrently, I was able to get about 25Mbps per download, totaling 50Mbps overall. Not bad!

Happy downloading!
See also at TinkerTry
Figure out what CDN your download is coming from:
-
How to download Windows 10 Technical Preview ISO Build 10162 quickly and easily
Mar 27 2015 - Using Task Manager and Resource Monitor to figure out what's going behind the scenes in Windows Server 2012 / Windows 8
Sep 07 2012
See also
VMware's official announcement:

All Comments on This Article (23)
Hmm, interesting, and thanks for reminding me of this little incident, now much more meaningful, given your report.
Here's the 7:50 spot:
https://youtu.be/VuJ8zqjwulg?t=7m50s
I'm working on figuring out if there's a valid/reasonable way of reporting such potential bugs to VMware, short of VMware Community posts that can sometimes get kind of lost, such as this pretty spectacular one:
https://communities.vmware.com/thread/547345?sr=stream
Will let everybody know if I ever figure that out (been working on that one for about 15 years ;-)
Paul, this is the same exact issue I had, at 7:50 in your video.
You say "There's something wrong with this VMFS, it's not being seen."
That's exactly what I experienced except it was post upgrading to 6.5 and I did indeed have VMs I cared about on my SM951.
Very frustrating and interesting watching you have the same issue. That's truly a nasty bug.
The drive would oddly on occasion show the contents right after boot of the host but as soon as it was accessed it would hang, and then disappear as per your experience in the video.
I haven't tested (unsupported) Veeam 9.5 on 6.5 yet, still plan too:
https://TinkerTry.com/veeam-availability-suite-9-5-nfr-code-available-with-vsphere-6-5-support-soon
Meanwhile, FYI, I did just recently test NAKIVO 6.2 with 6.5, and it seems to work just fine (haven't really tested it fully yet), easy to install and use (appliance), see:
https://TinkerTry.com/nakivo-6-2-nfr-available-for-download
This YouTube link should jump to 6 minutes 22 seconds in:
https://youtu.be/1qekrtveoFw?t=6m22s
to show the part you might want to briefly check out, where I install/configure, very simple, NFR (not time bomb).
You can speed my video up with this tip:
https://TinkerTry.com/video-speed-controller-chrome-extension-gives-playback-controls-to-any-html5-video
Hi Paul - thank you so much for doing all of this work looking into this and I'm terribly sorry for not replying to you sooner!
I do in fact have the AHCI version as I had pulled this from my Mac Pro (which isn't NVMe compatible.)
I have a SuperMicro X10DRi-T host and prior to the 6.5 update had never had an issue with the drive.
I do suspect that it was a 6.5 upgrade issue as I was able to format it fine (once I decided I didn't want to go through the trouble of trying to read a VMFS volume from macOS to recover the VMs!). One of these days I'll finally get around to setting up an NFR Veeam instance in my lab :)
Thanks again.
Hey "S", you're kinda famous, see your quote here:
https://TinkerTry.com/veeam-availability-suite-9-5-nfr-code-available-with-vsphere-6-5-support-soon
Sure wish your Disqus name said a little more about this mystery person who has helped in so many ways. Thanks again for all you do!
Perhaps this is only affecting people who upgraded to 6.5 (rather than my fresh install)? Not as crazy as it sounds, see:
https://TinkerTry.com/supermicro-xeond-superserver-slow-sata-driver-fix-for-esxi65#nov-17-2016-update-2
I even recorded my efforts to fail for you, right here:
https://youtu.be/VuJ8zqjwulg
I was not able to replicate your problem under a fresh/new install of ESXi 6.5 on my SYS-5028D-TN4T Xeon D-1541. I was on a strange mission, looking for trouble. Not sure if I should be happy.
I used freshly formatted VMFS 6 on a SM951 NVMe version. Is there a chance you have the AHCI version, and are somehow getting the slowdown Anthony reported?
https://TinkerTry.com/supermicro-xeond-superserver-slow-sata-driver-fix-for-esxi65
It's a long-shot I know, just trying to not leave stones unturned.
I couldn't replicate his results either, I'm failing to fail.
https://uploads.disquscdn.com/images/ea65e6c3862720519ab36f143ea3b15c37c45f47e9cde1e0529d70cb74e41758.png
Samsung 950 PRO seems to be working fine under ESXi 6.5 with the baked in driver.
https://uploads.disquscdn.com/images/1b95f9da879057d9d7cd881c60bb547138467edd76fef875fbe092e5f0945d7e.png
and for comparison, natively under Windows (2nd screenshot):
https://TinkerTry.com/samsung-950-pro-benchmark-results-superserver-sys-5028d-tn4t#results https://uploads.disquscdn.com/images/9707a8efdcae7d8c8e7984ab41e23356079f82daee9cbe8f60ba3dda5051faa0.png
That does sound interesting, and worrisome.
I have also have an 128GB (affordable) SM951 that I can try tonight (was used for my recent but now completed SYS-E200-8D testing). I have the NVMe version, not the AHCI version. Is that what you have, the NVMe version?
So far only brieflly worked with Samsung 950 PRO M.2 NVMe, which so far seems to be working as a VMFS datastore just fine, although this video doesn't quite cover that:
https://www.youtube.com/watch?v=DSJpAH93gcs
Just an FYI for Veeam users: Veeam doesn't support vSphere 6.5 yet (that includes Veeam 9.5 which was released today). From what I have read Veeam 9.5 Update 1 will support vSphere 6.5 and will only be available (roughly) in two months.
This is a bit of a show stopper for me as I use Veeam for my backups...
Reuben, I share that frustrating of not being able to report bugs (well, not without paying $299), and I even paid the $399 for the Academic vSphere Essentials AND EVALExperience AND have access to vExpert licenses. If you can shoot me an email at https://TinkerTry.com/contact#EmailForm I can include you on an email thread I have going with VMware right now actually. I even quoted you in that email ;-)
Anyone having issues with an M.2 SM951 SSD datastore post installation of 6.5?
It was working perfectly under 6.0 but as soon as 6.5 was installed it now hangs when accessed.
This is extremely useful information for other IPV6 users, so appreciative that you took the time to document this experience here!
Did you manage to format a 6.5 VMFS yet? 6.5's HTML5 web UI seems to default to 6.0 for new VMFS, will investigate tonight...
Your 6.0 keys _will_ work.
Yes I did originally, and in the process of putting all the IPv6 back on now post the upgrade. It's not looking good though, my VCSA accepts but then forgets it's IPv6 address and now I'm also getting some other useless nondescript error "Set DNS settings action failed due to: Operation Failed." I can no longer edit the DNS resolvers on the port 5480 admin page!
The migration/upgrade was pretty good once it got going, it basically does a clean install and then sucks the data out of the old one and into the new one.
But in so far as IPv6 goes - 6.0 was great, 6.5 looks really buggy-as-hell. Spin up a new VCSA and see if your experience is the same as mine.
I also am rather stuck with the Nexus1000v. The VIB won't install on account of a dependency on a 6.0.0 base vib, not sure how I am going to get around that one just yet.
I suppose this is the price for being a very early adopter. Kinda sucks that I can't even log a ticket with VMware to get them to fix -their- bugs though.
Thats worrying as I run IPv6 on everything at home including all VMware products (where I can).
Do you have DNS setup/configured for IPv6?
Turns out the problem was to do with IPv6 and I've now gotten a migration to work succesfully. I have had a dual-stacked VCSA for some time now, but this created havoc with the migration tool.
To fix this problem I had to remove all traces of IPv6 on the VCSA, and then disable IPv6 entirely on the console, restart, and then go in to the bash shell via SSH and remove all of the IPv6 configuration that remained on the interface-specific network configuration in /etc. The console will stiff bitch and complain about invalid IPv6 configuration even if IPv6 is disabled.
It seems like there are a lot of bugs around this process, the migration GUI doesn't offer options to configure both IPv4 and IPv6 as part of the migration and it dies if you try and migrate a dual-stacked VCSA. And then when you do remove the config in the GUI it doesn't actually remove all of the IPv6, which of course causes the migration to fail mid way through after it starts pulling data off the old VCSA and putting it on the new one.
I did sign up for the beta, but, well I was never asked to participate, so........
Not so good here. I have been attempting to do an upgrade of VCSA 6.0 to 6.5 but it fails with an error:
Error setting DNS configuration. Details : Operation Failed.. Code: com.vmware.applmgmt.err_operation_failed
Failed to set DNS
Even though I only have IPv4 configured, and one single DNS server set. I've tried it twice. The logs indicate that the system is trying this:
2016-11-16T02:09:30.694714+00:00 localhost vm-support: Failed to run cmd: reg query HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\VMwareAfdService /s: [Errno 2] No such file or directory
Which makes absolutely no sense whatsoever on a Linux based VCSA.
I'd be interested to hear if anyone else has had success. It's pretty painful to create a new VCSA with new dvSwitches, remap ports, re-attach Nexus 1kv etc
Allan
How exactly did you all the traces of ipv6? I have the same problem.