Strange Error after upgrading to v6

[mroof]

Hi Everyone,

Things were working just perfectly with v5 but now after my upgrade I'm getting a strange error on about 1/3 of VMs. "Error: An error occurred while parsing EntityName. Line 1, position 8." I can't seem to figure out what in the heck is causing this one and I haven't heard back from support (ID#5159844) yet so I was hoping someone on the forum just might know what is going on.

Thanks!

[Gostev]

Hi, I can confirm that this has not been reported yet. But I can suspect this to be an upgrade-related bug. Thanks.

[mroof]

I saw another post about this on when I Googled it in the Veeam forum but now when you click on it the topic is gone. Any idea what happened to the other posts?

[Gostev]

It must have been hit by one of 4 bullets from Missing your post? sticky topic. By the way, please edit your original post to include support case ID before more evil moderator shows up

[mroof]

After working on this issue some more it looks like this is related to how Veeam skips swap space used inside the virtual machine. I noticed that if I watched the trouble VM while Veeam is trying to backup it up it always stops sort of backing up the entire HDD and then when it lists the swap space it skipped the totals don't match. However, on the windows VM's where it does work the totals do match correctly for the space used.

So, far I have removed the pagefiles with the system control panel and then recreated them but at the moment it hasn't helped. I will keep trying on finding a solution. If anyone else finds one before please post it.

[Gostev]

Don't dig towards Windows page file - to the best I know, the error above is clearly unrelated. Try to create new job and see if it helps.

[mroof]

Did that already. Tried to create a new only for one troubled VM and I get the same error.

[Gostev]

OK, then most likely the issue sits in actual configuration database that was brought over from v5...

[mroof]

Okay, I found the problem!

For me it was because I had a datastore whose name had a "&" in it. Ie, "Disk & Other data Storage". When Veeam would backup the disk it would write everything and then make an xml call. The string it would try to store looked like "[Disk & Other...." and position 8 is just after the "&". So, I bet that even though VMWare escapes such characters just fine there is a spot in Veeam where they are not checked.

For others that are getting a similar issue try removing any special characters you might have used in the names of datastores (maybe even VM names) and see if it goes away. If does please report it so that support can let the engineers know at some point so it gets fixed.

[Vitaliy S.]

Hi Morey, non-latin characters are not supported in the product’s installation path; in the backup target path and file names etc. as per our Release Notes document.

Thank you for posting the resolution!

[Gostev]

Yes, but Morey is talking about datastore name. I think this is a bug.

[mchapman]

I concur we changed our datastore name to NOT include the ampersand and the jobs we were experiencing the issue with completed without an issue