Difference between replication with seed and without seed

[baber]

dear all
Hi

i have a question about replication in veeam :

i am realy confused

we can do replication with 2 method in veeam :

1 - without any change in backup and replication tab right click and create a replication job without check enable replica seeding
it will do replication without any problem in first time will be do a full sync after that do replication and just move data changes


now we have other method that if want use replication job with check enable replica seeding in this method we have to first backup from source vm and then copy that on the repository of DR site and will be do replication everytime such as previous method now i want to know realy there is not many difference between two methods the time for second replication in both of method was same in second method we check enable replica seeding and in first mwthod we don't have this option but actuarly in first method it will automatically use from seed because it don't move all of data in each replication

now i realy confused what is difference between two method of replication :

replication without seed and replication with seed

BR

[DGrinev]

Hi,

You can have different reasons to seed a VM for replication, such as expensive traffic cost between sites or slow WAN connection.
By seeding a VM you will significantly reduce the amount of data to be transferred on the initial run of replication. Thanks!

[baber]

thanks

but in previous post i say that in replication without seed for do replicate after first replication i think it will use from seeds because it don't full replication just send changes and with seed also it will be send just changes now what is difference???

[DGrinev]

As you already mentioned, on the first run without seed the replication job will move Full VM data and then only changes during following runs.
With enabled replica seeding the job will compare the differences between data blocks of the source VM and the seed to initiate transfer only for changes. Thanks!

[baber]

but in both mode in second replication it just move changed data not all of data
is that difference???

[DGrinev]

Correct. The difference is only in the 1st job pass.

[sasam]

I believe this question was more in a way of, Why would anyone have to choose between replication over network without seed and replication over network with seed (by copying this initial seed over the network)

The answer is that you do not have to do initial seed by copying that initial backup over the network.
You can copy it to USB drive or some other unit and physically move it to secondary location and use those files as seed.

The one thing that is not easily mentioned is that, upon first run of replication with initial seed (when you set it up and everything starts running), that initial replication job will run for a long time, even though you have moved data already to secondary site.

It was not clear for me at first why, but working with my network engineer, he noticed that not much data is transferred from site A to site B during this, so I figured it is only checking blocks to see what is different between that "initial seed" and your now working machine, as it does not have information on changed blocks at this time.
Next time repliciation runs, it only transfers changed blocks as it has information where they are and only transfer them.

If I said something that is not correct, someone from Veeam can add to my post here, but this is how I see it.
Thanks
S

[DGrinev]

Hi Sasa,

Your understanding is correct.
The main purpose of the replica seeding, as I already mentioned is to reduce amount of traffic that can be huge due to size of VM.
Sometimes it's much cheaper to send USB drive over delivery service or bring it yourself than transfer GBs/TBs through expensive or slow WAN. Thanks!