Re: Unanticipated time ... for full recovery

[hvjunk <hvjunk@gmail.com>]

> On 13 Aug 2025, at 10:17, Tomaž Šolc via tarsnap-users <tarsnap-users@tarsnap.com> wrote:
> 
> I'm not sure if these download rates are a technical limitation or 

The problem with restores in the design of tarsnap: 

 Each block has to be fetched, and then the next block could be requested. In that is a latency, and the TCP/IP etc. get cold and never ramps up properly.

This gets worse the further away (network latency ping) you are from AWS EAST-1 West Viginia (If I recall correctly). There are no CloudFront to assist etc. as everything is run on the AWS server in AWS EAST1.

It had been a complaint since pre-2020 from me (trying to restore to France based servers.. and even the Canadian servers had a similar, but lesser) as ANY latency just makes tarsnap restores unbearable to say the least. 

The fastest solution: spin up a VM in AWS EAST1 (or wherever tarsnap server is located) restore to there, and then transfer to where ever you need.

I’m sorry, but that is the reality of tarsnap only supporting the one AWS region/zone. 

I’ve since moved my backups to OVH cloudarchives or ProxMox BackupServer with synchronization to providers like Tuxis.nl for this reason.

I like the product’s design etc. but it just doesn’t cope with latency  ;(


Hendrik Visage