New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Have you tried a different network, like not M247?
thanks for the comment, why did you add M247?
You likely see TCP limitation caused by RTT.
You have ~110ms ping between these locations so ~3Mbps is completly normal for TCP.
Divide backups in parts and speeds will multiply.
I get around 100 Mbit/s. Not phenomenal, but not too bad either.
M247 is famous for their congested network and HostHatch is famous for "fixing" routing issues.
ie 288KB/s is ok for this? I'm just seeing this for the first time
do you think this is the reason? I backup there from all over the world, more than 10 hosts, no problems
I have to say I use HostHatch Amsterdam location for my backups its a super fast location over 2 years everything is ok, but today i tried to send a file from the server to Serverica and I was shocked by the wait time
>
This is dumbest thing I've read this month.
I tried again 487KB/s eta 3d 14h
Redacted by Moderator.
@angstrom @Arkas @DP
The jump from this
to
made me giggle its just so unexpected lol.
how are you copying? public IP to public IP ? trough a tunnel (vpn?)
oh and the bandwidth vs latency is BS..
@bench, such comments are not welcome here.
Take this as a warning.
Cheers.
Dumb Cisco, they dont know shit about networking
https://accedian.com/blog/measuring-network-performance-latency-throughput-packet-loss/
"TCP throughput with no packet loss Round trip latency
90ms = 5.32 Mbps"
Thanks
Try to transfer two files in parallel.
If bandwidth doubled then you are limited by waiting for ACK - server stops sending more data until client responds tgat earlier packets are okay.
You'll see the same effect in iperf3.
UDP is not affected by it, only TCP. I do not know what you are using for transfering backups, but like I said - for TCP single stream its normal to get these transfers at such high RTT. Its fixeable by just chunking backups, like most backup software does, because then you have multiple transfers at once.
I do this all the time, today there is some kind of anomaly, thanks for the advice, maybe there is a problem on both sides or even not on the one I thought about at first
There may be also packet loss somewhere, check MTR between these two locations. This would also explain why your results vary so much
whatever the problem is now, even the theoretical one i was warned about, it's a problem, my file will be in Canada in more than 24 hours and that's sad
welcome to the Internet
Can confirm!
Maybe try enabling TCP BBR assuming this is a TCP based protocol.
What was the redacted comment?
Something Russian, Putin, cutting subsea cables, insert swear word here multiple times.
Lmao that’s just funny
I am happy with the througput on my storage plan from servarica. These numbers are between servarica and a residential 1gbps fiber residential service in Central USA.
If i utiize a wireguard connection it drops to about 625 mbps
If I run iperf3 in Reverse I am capped by my residential upload speed of 100mbps
mtr home # 11 hops ~42 ms ping
-- another example ---
from one vps to another:
between servarica and RN in atlanta can get 500-600 mbps either way via 13 hops ~32ms ping