Well--I used to be able to export (not talking about OVA, but rather OVF)--with 6.5, but since going to 6.7, I now have the following problem:
- ESXi goes through the motions of exporting well-enough--meaning it makes the attempt. But, if I use Chrome to accomplish (actually I am using Canary), I almost immediately receive a network failure message--something that if I Google that problem, has existed for quite some time, now, and I am incredulous it still exists.... Google needs to get is act together and fix their browser instead of invading everyone's privacy
- If I use Firefox (which used to work when I realized Canary was routinely failing with a network error); what I see when I used Firefox, is the actual OVF file (config file) gets created ok, but any disk (vmdk) file the VM possesses, does get created (which is more than what Canary was doing for me), BUT, the disk file is way smaller than it should be
Let's say, for example, I have a 63GB Disk0 (the actual disk is "larger"--what I really mean is this is how much space the VM is actually using according to ESXi Storage, the datastore browser and drilling down until I actually "see" the (disk) file there.... When I export this in Firefox, I might get 57MB during one export session, and after I delete the files Export created and command the operation again, disk0 might show up as 23MB in another export session. I should have gotten something measuring in the GIGABYTEs--not MEGABYTEs.
When I tried exporting 2 of my VMs today, I tried multiple times and wrote down what Firefox saved for the four files created (1 ovf file and 3 vmdk files)--during each session. The ONLY file consistently created the same size each time was the actual .ovf file. The 3 vmdk files were a different size each time--and nowhere near the size they should have been. Of course, ESXi happily reported the export completed satisfactorily, even though nothing could have been farther from the truth. And while I did not bother trying to import the files created, I know ESXi would have said they were corrupted or otherwise unusable. I know this as someone elsewhere in the Ether reported this very thing.
When I compare successful exports I performed last year under ESXi 6.5 (using earlier versions of Firefox, too), typically disk sizes (depending on the VM I exported) ranged anywhere from 14 to 60GB for disk0. No way was 57MB anywhere close to what should have been created for disk0.
As others have indicated, this needs to be "sorted out". I suppose it is "nice" that OVF Tool works, alternatively, but to me, why must I continue learning myriad workarounds for something that I know used to previously work just fine. I've never needed OVF Tool and never had to use it (and honestly I don't know how). I've never needed vCenter, either, as I don't manage a huge server farm. Using the embedded client is all I really need. At first, I bemoaned the loss of the C# client (and in times like this, I still do, as that tool flat-out worked), whereas the web has, (how shall I say it) "quirks".
And speaking of quirks, here's a few others I've seen, incidentally:
- The reported addition of the copy and paste isolation tool parameters doesn't seem to work--In fact I have NEVER gotten copy and paste (shared clipboard) to work. Ever--regardless of VMware version. To make it work, I have to physically RDP into the VM rather than use anything VMware-oriented (such as VMRC). Only then does copy and paste work.
- Even though I have set the timeout to 0 in the embedded client, I have encountered unexplained lapses in the web's ability to respond/display things--forcing me to log out and log back into the embedded web client, restoring normal behavior--at least for a while, when I have to repeat the logout/login again. This has been especially true since upgrading to 6.7.
Of the above items, however, I consider not being able to successfully export a VM, a (very) major matter--especially since this is part-and-parcel a rudimentary function of ESXi. I can appreciate how "difficult" this might be to troubleshoot, but at my locale, it is not happening randomly. It is quite repeatable....
If there are any logs or such I could provide to help someone fix this, tell me where to look, and tell me where to send whatever I find....
Best Regards....
