Remote Replica from Backup performance and Per VM Backup Fil

[swiniarz]

Hello,

We are setting up Remote Replica from Backup and noticed some strange / poor performances :

- For each VM processed, the job consumes something like 20 minutes at source side before sending data over WAN to target ESXi DS. This happen for each VM and so it take a long time to replicate all our VMs (138).
For the moment improving the number of Parallel Task is the only way i find to reduce replication window.

Can someone know what Remote Replica from Backup do during those 20 first minutes ?

Another question about this problem, does using Per VM Backup file instead of traditionnal backup at source would speed up replication ? I've read that Per VM Backup file produce only one MetaData file so if it is the parsing of this MetaData that slows replication, Per VM Backup file would not help...

Regards

[veremin]

Can you double click on a job, select a VM, and check what particular step took more time than expected? Thanks.

[foggy]

Specifically, what operation is logged at that time in the job session log.

Per-VM backup chain on source should not affect performance of the replication job.

[swiniarz]

Hello All,

When i deploy one replication job then select a VM a I got some information about duration but nothing that show a process that last a long time ...

I've looked in one VM replication job and I see an important number of failed connection that delay next try for about 30s ...

Code: Select all

Failed to connect to agent's endpoint '172.18.0.12:2507'. Host: 'swvea012p0.aveo.local'.
[11.07.2016 10:21:07] <18> Error    Une tentative de connexion a échoué car le parti connecté n’a pas répondu convenablement au-delà d’une certaine durée ou une connexion établie a échoué car l’hôte de connexion n’a pas répondu 172.18.0.12:2507 (System.Net.Sockets.SocketException)
[11.07.2016 10:21:07] <18> Error       à System.Net.Sockets.Socket.DoConnect(EndPoint endPointSnapshot, SocketAddress socketAddress)
[11.07.2016 10:21:07] <18> Error       à System.Net.Sockets.Socket.Connect(EndPoint remoteEP)
[11.07.2016 10:21:07] <18> Error       à Veeam.Backup.AgentProvider.CAgentEndpointConnecter.ConnectToAgentEndpoint(ISocket socket, IAgentEndPoint endPoint)

[foggy]

Vladimir and I were talking about job statistics window available in the job's shortcut menu. At what operation the job is stuck during those 20 minutes? I also recommend contacting technical support with this issue for more effective troubleshooting.

[veremin]

Can you also verify the connection to the given host? Isn't there anything in between that might result in failed connection such as firewall, etc.? Thanks.

[swiniarz]

Ok, i'll check at next run of replication job .
Ok, we will also open a case.

Thanks

[swiniarz]

@ v.Eremin

Network connection between sites work fine but it sometimes take some time to estabish (aka RDp, ...) , may be an hint ...
Is there a way to change connection timers during a replication job ?

regards

[foggy]

There are some configurable timeouts, support should be able to review logs and define whether any of them are applicable.