Failed to create NFC download stream

[SimonEa]

...

I solved it by adding an entry for the ESXi Host to the hosts file of the backup server (with veeam running on).

This also worked for me

[michaelwright1971]

Hya!

Im getting a similar issue when trying to perform a direct SAN access backup after copying the vm files to another ESXI host in DR

the vm is up and running but when I try to perform a backup I get:

Code: Select all

06/11/2013 18:27:39 :: Processing 'prod02-Veeam' Error: Client error: DiskLib error: [16].No file exists for given path
Failed to create NFC download stream. NFC path: [nfc://conn:10.222.1.95,nfchost:ha-host,stg:521f26b0-9c4e151d-2415-d89d67243d94@[DB-Raid-10] prod02-Veeam.vmxf].
Failed to backup file. File: [nfc://conn:10.222.1.95,nfchost:ha-host,stg:521f26b0-9c4e151d-2415-d89d67243d94@[DB-Raid-10] prod02-Veeam.vmxf]. ItemSpec: [veeamfs:5:8b823c4a-24d7-4a2c-815e-f408e23fffd5 (28)\[DB-Raid-10] prod02-Veeam.vmxf@/ext-nfs

[Gostev]

Hi Michael, if you need help with this error, then please open a support case with the complete log package, and post the support case ID here. Thanks!

[michaelwright1971]

Hya Gostev,

I have created a case and the ID is: 00473625

Any help would be appreciated..

cheers Mike

[michaelwright1971]

I have now configured the backup job through vcentre and im getting:

Code: Select all

06/11/2013 20:08:20 :: Processing 'prod02-Veeam' Error: Client error: DiskLib error: [16].No file exists for given path
Failed to create NFC download stream. NFC path: [nfc://conn:10.252.1.99,nfchost:host-20,stg:datastore-22@[DB-Raid-10] prod02-Veeam.vmxf].
Failed to backup file. File: [nfc://conn:10.252.1.99,nfchost:host-20,stg:datastore-22@[DB-Raid-10] prod02-Veeam.vmxf]. ItemSpec: [veeamfs:5:eeb55650-20cc-4875-a254-23a385e6d5f6 (vm-121)\[DB-Raid-10] prod02-Veeam.vmxf@/ext-nfs2/Backup Job 3/Backup Job 32013-11-06T2008

cheers Mike

[tsightler]

Can your proxy reach the ESXi host via the IP address 10.252.1.99?

[michaelwright1971]

There is one ESXI host 10.252.1.95 which controlled by the vcenter with the address 10.252.1.99(which is a vm on the host 10.252.1.95)

the proxy im using 10.252.1.99 as shown above is also the vcenter

cheers Mike

[Berniebgf]

Hi All,

Have a poor environment configuration (not my doing) that I am trying to get Veeam replication working in and am having problems.

Issue is: New replication jobs (never had replication in this environment before) SOAP Failed to Create NFC download Stream

Environment is as follows.

1. VC Server & Veeam, Server installed together - 100 subnet
1.1 Lots of network interfaces on this server connecting to 100 / 110 subnets.
2. 2 x ESXi 4.1 registered to the above virtual centre server - their management ports are in 40 subnet and 60 subnet (I know.....god help me).
3. Juniper firewall is acting as a router between the subnets.
2 x new proxies in 100 subnet

Replication to the following environment (Layer 2 connection) (we setup this new environment)
1. another virtualcentre server in 100 subnet
2. 2 x ESX hosts with management interaces in 100 subnet
3. 2 x VIrtualised proxies in 100 subnet

The backups for this environment work OK, rules were put in the firewall to allow communication...

Replication does not work, I am selecting a source proxy (100 subnet) and a dest proxy (100 subnet)....... they can ping each other and resolve DNS.

I have check that Port 902 is Open between the prod proxy and the virtualcentre / esx management ports
I have check DNS resolution.
I have tried different proxies confirming they have access to the resources needed...

Anyone else have idea's?

Bernie

[Berniebgf]

Update after more research.

Can across the following link - http://forums.veeam.com/viewtopic.php?f ... 1&start=15

It has led me to the problem which is the destination site of the replication virtualcentre server not the source.

"Soap fault. No DataDetail: 'get host by name failed in tcp_connect()', endpoint: 'https://XXXXXXXXXXXXXXX.com.au:443/sdk'"

I now know it looks to be a DNS issue with resolving the destination side VirtualCentre server.

Question now is who is trying to talk to who? destination proxy to destination VIrtualCenter? Source Proxy to Destination VirtualCenter? ...... I'll work it out...

Bernie

[Berniebgf]

Another update...

Have opened call with Veeam.

Check both source and destination proxy logs.... getting the following on BOTH as they try to create NFC connection with source and destination VC servers.

Client error: Cannot get service content.
[12.11.2013 11:50:38] <11> Error Soap fault. TimeoutDetail: 'connect failed in tcp_connect()', endpoint: 'https://XXXXX:443/sdk'
[12.11.2013 11:50:38] <11> Error SOAP connection is not available. Connection ID: [localhost].
[12.11.2013 11:50:38] <11> Error Failed to create NFC download stream. NFC path: [nfc://conn:XXX,nfchost:host-145,stg:datastore-150@XXX/XXX.vmx].

Maybe its an Authentication issue as non of these servers or proxies are on a domain (Not my doing).

Bernie.

[ClausS]

I moved my VM files betwen different virtual iSCSI Drives (Windows Server 2012), which I mounted on a free ESXi 5.5 system.
Suddenly I run into these error messages with VEEAM 7:
... No file exists for given path
Failed to create NFC download stream. ...

I tried restarting the ESX, removed all machines from VEEAM 7, added them again and had the same problem.
I rebootet the VEEAM system, the Win 2012 fileserver and the ESX, but the problem was still there.
I also created hosts-files entries for all systems and checked DNS, but the problem was still there.

Then I noticed that one machine of the ESX was backuped properly. The difference between this "good" VM and all the
others was an addtional folder in the filesystem.

The "good"configuration: iSCSI mountpoint / {VM name} / {VM files}
"Bad" configuration: iSCSI mountpoint / {VM files}

Now I removed the "bad" VMs from vSphere client list, moved the files (via SSH) to a subfolder and imported the
*.vmx files from their new location to the vSphere client list.

=> Problem solved.

[Vitaliy S.]

Claus, thanks for your time spent giving us feedback and your resolution on this issue. I wonder how can we reproduce "bad" VM configuration path...

[ClausS]

Claus, thanks for your time spent giving us feedback and your resolution on this issue. I wonder how can we reproduce "bad" VM configuration path...

Good question. What I did:
I had several VMs on local HDDs in another ESX box. Just single disks, no RAID. These VMs should be transfered to a more secure storage. I played arround with the new storage spaces on a Windows 2012 R2 Server. The result was a 2-tiered mirrored storage as iSCSI target. Next I restored some of the VMs with VEEAM from backup to this iSCSI targets. That works fine, but was slow (20MB/s). I think all VMs that have been restored/moved with VEEAM are fine to be backuped again with VEEAM.
For the really big VMs (~1TB) I used another way: I mounted the virtual iSCSI drive to the source ESX and copied all files of the VM in a SSH shell to the their new location on the root folder (= "bad" VM configuration path) of the iSCSI drive (each VM has it's own virt. iSCSI drive). I think that causes the errors later when doing a VEEAM backup. The VM itself runs properly with this configuration.
I searched the *.vmx files for wrong filesystem locations, but could not find any. But I'm not very familar with working at a shell prompt with vi ...

Normally people will not run into this problem, if everything is done by VEEAM.
But until today I thought I was also normal too ...

[Vitaliy S.]

Yes, seems like this operation might lead to invalid VM path configuration, thanks for sharing!