BitTorrent Performance Test: Sync Is Faster Than Google Drive, OneDrive, Dropbox 124
An anonymous reader writes Now that its file synchronization tool has received a few updates, BitTorrent is going on the offensive against cloud-based storage services by showing off just how fast BitTorrent Sync can be. More specifically, the company conducted a test that shows Sync destroys Google Drive, Microsoft's OneDrive, and Dropbox. The company transferred a 1.36 GB MP4 video clip between two Apple MacBook Pros using two Apple Thunderbolt to Gigabit Ethernet Adapters, the Time.gov site as a real-time clock, and the Internet connection at its headquarters (1 Gbps up/down). The timer started when the file transfer was initiated and then stopped once the file was fully synced and downloaded onto the receiving machine. Sync performed 8x faster than Google Drive, 11x faster than OneDrive, and 16x faster than Dropbox.
and speed was never the point of dropbox (Score:5, Insightful)
Re: (Score:3, Informative)
That's completely independent of speed, what you are talking about are limits and/or throttling, both of which are sliders in settings.
Re: (Score:2)
...
And who knows, I didnt care to read tfa, but they couldve developed a nice algorithm like that new hit show I cant remember the name of
Re: (Score:1)
...and like that new hit show, such an algorithm will ever remain the figment of someone's deranged imagination.
(I mean, seriously?? A plane that size doing 200mph eight feet off the ground not a: smashing engine pods to bits and/or b: flipping due to ground effect? And why couldn't they just fucking land it, they had the time and apparently the runway...)
Re: (Score:1)
"(I mean, seriously?? A plane that size doing 200mph eight feet off the ground not a: smashing engine pods to bits and/or b: flipping due to ground effect? And why couldn't they just fucking land it, they had the time and apparently the runway...)"
And that other show, where there is that doctor that cannot die, who, when killed, appears under water, naked.
But apparently the phone in his clothes, that disappear with him, are magically transferred to a new suit at home, where he can use it again when needed.
I
Re: (Score:2)
Actually I'm pretty sure he never carries a cell phone. In the last episode one rings in his pocket, having been planted there, he's surprised, and he ultimately ditches it. One must wonder about his wallet though... Overall Forever hasn't been all that bad. Scorpion has also gotten a little less far-fetched but not much. It's still fun to suspend belief and just watch rather than trying to pick it apart.
Re: (Score:2)
nah, it's still more fun to poke holes in it...
Re: (Score:2)
And that other show, where there is that doctor that cannot die, who, when killed, appears under water, naked.
Wait, what show is that?
Or was it a dream you had?
Re: (Score:2)
Forever (2014)
http://www.imdb.com/title/tt34... [imdb.com]
It's still finding itself but it's actually doing a pretty good job of it so far.
Re: (Score:1)
Silicon Valley? "Meet in the Middle" / "Pied Piper"?
Re: (Score:2)
Black Jesus?
Re: (Score:2)
Re: (Score:2)
And all the others were uploading an online copy too, so this isn't a like-for-like comparison (I think that DropBox, at least, supports direct local network copying but would probably be making the online backup at the same time).
Re: (Score:2)
You should be relying upon bandwidth throttling features to do that for you, not the inefficiency of your tech. Speed is vital in this market because speed reduces the chances of causing data conflicts. Slow and steady background uploads increase risk of conflicts as the average user doesn't pay attention to upload/download status before shutting down a machine or resuming work. Faster transfer reduce problematic "lazy" sync usage at a statistical scale.
Is it open source yet? (Score:4, Insightful)
There's no real point in using it if you can't even trust it does what they say it does...
Re:Is it open source yet? (Score:5, Informative)
Here's an 'unofficial' open-source bit-sync client:
www.yeasoft.com/site/projects:btsync-deb:btsync-server
It doesn't install on .rpm based distros so far as I can tell. I have a use-case that calls for drop-dead-easy cross-platform sync, and I'm leaning towards using git-annex assistant [branchable.com], but haven't had time to thoroughly test it yet.
That isn't open source (Score:2)
That isn't an open source implementation of btsync. It is just an unofficial debian package that installs the official proprietary btsync binary. It makes it easier to install and update btsync on debian based systems, but it is the exact same software that you download from the official site.
Re: (Score:2)
No, but there is an open source bittorrent based syncer called "syncthing". It's not as mature, but it is supposedly functional. Have at :)
Re: (Score:2)
Re: (Score:2)
For my workgroup/use-case we looked into Syncthing/Pulse and ruled it out because another requirement of ours is read-only sharing/distribution. So far, we're still stuck using the official Bit-Torrent Sync. In other words, so far, given our long list of fairly strict requirements, Bit-Torrent Sync sucks less then everything else.
I'm hoping git annex assistant will pass testing once I get to it. We're trying to distribute files using wifi at open-source conferences, using some kind of LAN technology, since
Re: (Score:3)
None of the services listed are open source, so that is a red herring. Open source isn't even particularly important here, because your data isn't locked into any kind of a format - you can switch freely to any service at any time, and you have a complete copy of your data at all times. If you really need open source, there are options which require a server: SparkleShare works well for me, and I understand that OwnCloud has something that works decently as well.
My problem with the service is that it works
Re: (Score:2)
OwnCloud is open source and does the same things as Dropbox (although in really crappy PHP on the server, so you'd better have a lot of spare cycles to burn - it's the first time for several years I've seen file transfers across the Internet be CPU limited).
The problem is that they're comparing apples to oranges. Of course a direct local connection will be faster than two devices sharing the same Internet connection and going via a server, but most of the time that I want to use a server as part of a sy
Re: (Score:2)
It doesn't even do anything for me. It was working and then just stopped. No files synch for me and both the handset and the computer it's supposed to synch with are 'net reachable. Nothing changed in the config of the PC so I have no idea what happened. I've tried dorking with settings and updating every time something new comes out - doesn't help...
Re: (Score:1)
Just hit the menu button and it will give you an option for "Sync".
It might start working again after that, also there is a possibility that the options for your folders has changed. They have the option of syncing automatically or just syncing the file description until you click on it.
I hope this helps you.
Re: (Score:3)
Google Drive, OneDrive, Dropbox
They all have your data, they can do whatever the f... they want with it. Unless you're talking about a client backdoor to access all the other files you didn't want to share with the cloud, but I don't think any of the others are any better. If you want real control, it's ownCloud or no cloud I think...
Re: (Score:3)
I've been meaning to ask someone about this. Is OwnCloud something that someone who's kind of a moron could set up on their own server? Asking for a friend.
Maybe not a moron, I mean, I've set up Apache and a media server, and I can read instructions when I'm sober. I just worry that I'll do something wrong and end up syncing my data with some Estonian hackers by mistake.
Re: (Score:1)
If you've managed to set up an Apache server, you've got enough prerequisite skill to be able to install OwnCloud. I set it up once and it works, kinda, but it's not really as slick or robust as something like Dropbox. It's still a fairly immature FOSS product with not much in the way of paid development and it definitely shows at times in terms of usability and performance.
It's great fun to roll your own at times, but you'll always be behind the pros who do this for a living and have worked out all the kin
Re: (Score:3)
Supposedly it has gotten a lot better in recent versions:
http://owncloud.org/blog/owncl... [owncloud.org]
Re:Is it open source yet? (Score:4, Interesting)
It's not that difficult. But after setting it up for a group of people and then setting up seafile, I prefer seafile. If you aren't an admin user in owncloud, things are pretty tough when it comes to knowing what groups you are in and what groups can be shared with and such. seafile does a much better job on that front.
Plus the owncloud sync client doesn't seem very good. And the mobile platform clients cost money where seafile is free.
ownCloud might have gotten the 'good name', but they don't have the best implementation sadly.
Re: (Score:3)
Seafile. Got it. Thank you.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Re: (Score:1)
de-compile the code? In Open Source the code is there, and people do look at it.
Re: (Score:2)
Is it open source yet? There's no real point in using it if you can't even trust it does what they say it does...
I can trust a company even if a program is not open source.
Re: (Score:2)
This may be the most /. ever example of the bravery/stupidity conundrum.
OwnCloud? (Score:1)
Re: (Score:3)
If off-site from the server, I doubt the OwnCloud clients are smart enough to know a friendly computer is on the same LAN to share already downloaded chunks.
Which I might add is the only advantage to Bittorrent Sync. The technology only provides an increase in speed if one of the clients on the LAN has pieces of data already downloaded so the Internet connection is not as necessary. If neither comput
Re: (Score:2)
I'm not certain, but it's quite possible that the server they're both downloading from is smart enough to say "Hmm. I have two clients who want the same thing and have none of it. Let's try sending them different blocks and see if they can share between themselves; if they can, I only have to send each block once instead of twice."
Re: (Score:2)
(Speaking of BTSync, I was. That was a standard optimization for seeding, to try to get swarm members to get stuff from each other instead of slamming the seed.)
Re: (Score:2)
Re: (Score:2)
Just like Bittorrent sync, its highly dependent on your setup. If you run Owncloud on your home router with 1M uplink, your speed is that small. If you run your owncloud on a server with a gigabit uplink, and use google fiber, and you have an SSD in your owncloud server, you might get faster speeds.
Re: (Score:3)
OwnCloud is a WebDAV based system. It's inherently bloated, but it works. Setting it up your own web server is a requirement (or purchasing web hosting somewhere, but then the trust/security goes out the window).
Google Drive, Dropbox, Onedrive, OwnCloud all require storing your data elsewhere.
BT Sync only syncs data across your devices. It does it really well, utilizing Bittorrent protocols and DHT. It's actually a very useful tool. I use it all the time.
Comparing LAN to WAN Speeds (Score:2, Insightful)
In other news, you spend less time on an airplane when you take a staycation.
Re: (Score:2, Interesting)
Actually, while they indeed compared two computers on the same LAN, they also included a computer on the internet. Furthermore, One of Dropbox's touted features is that it's able to detect and use peers on a LAN to avoid the unneecssary round trip through the cloud. I don't know about Google Drive, but judging by the results I suspect they can do the same.
And, more importantly, they compared the other clients on the same setup.
How you got modded "+4 insightful" is beyond me.
Re:Comparing LAN to WAN Speeds (Score:5, Informative)
"Our tests were conducted over local LAN – on the same switch – in order to rule out available bandwidth as a limiting factor. It’s important here to note that Dropbox, Google Drive and Microsoft OneDrive all rate-limit uploads and do not fully utilize the 1 Gbps bandwidth available (in regards to the office Internet connection, not the LAN switched). We’re confident that a slower Internet connection would yield similar results."
In other words, people agreed with me because they knew what I said to be true.
Not only did they give themselves the preferential treatment of same LAN, they also intentionally adjusted their tests to discount an advantage of a competitor. Again, quoted verbatum from the blog post:
"Dropbox has a deduplication scheme in place – what this meant for our tests is that even though we deleted the video file from our Dropbox folder, traces of it still remained and Dropbox got ~50% faster at transferring the same video file each subsequent time we uploaded it. To correct for this, we needed a new file that wasn’t bit-for-bit identical to the video file we previously transferred. "
Why don't you RTFA.
http://blog.bittorrent.com/201... [bittorrent.com]
Re: (Score:2)
They compared the transfers between two laptops on the same LAN
a) we don't know whether the two laptops could talk to each other across the LAN - in fact without evidence to the contrary I'd assume they couldn't
b) Dropbox will sync across the LAN if it can.
In any case, I'm not sure the LAN/WAN distinction is too relevant here, given that they were using 1gb/s internet connection.
Re: (Score:2)
b.) True, and that is why they intentionally nerfed the Dropbox test by creating a new random file to not only avoid "deduplication" as they say, but the LAN Sync being available as well (which they do not admit).
RTFA:
http://blog.bittorrent.com/201... [bittorrent.com]
Re: (Score:2)
a.) Yes, we do because the blog post says as such.
I don't think it's at all clear.
The company transferred a 1.36 GB MP4 video clip between two Apple MacBook Pros using [...] the Internet connection at its headquarters.
Sync’s time might seem ridiculously low, almost as if the Internet wasn’t involved at all. You have to remember, however, that BitTorrent’s headquarters has a ridiculous fast connection both downstream and upstream.
That implies that the LAN wasn't involved, although it's not very clearly stated.
In any case the blogger repeated the test over the internet and got similar results.
Trickle (Score:4, Informative)
These programs are designed not to saturate the upload/download pipes ruining the connection for all the users. So congrats, your protocol has all the problems of BitTorrent.
Ruining the connections since 2001.
Re: (Score:2, Redundant)
Well, quite:
More specifically, the company conducted a test that shows Sync destroys Google Drive, Microsoft’s OneDrive, and Dropbox.
And your internet connection at the same time.
Re:Trickle (Score:5, Informative)
Re:Trickle (Score:5, Informative)
The reason loss occurs after the clamping is because my ISP uses small buffers. They don't like buffer bloat. My max latency to my ISP before loss starts to occur is about 10ms. Since the connection is dedicated, and their trunk is sized about 3x more than peak bandwidth, it's normally not an issue.
This wouldn't be an issue if DropBox transferred the data as a single stream, but instead does a very jarring start/stop cycle, which causes my bandwidth to get very spiky. I'm thinking of enabling traffic shaping on my PFSense box, but I really don't feel like messing with it quite yet.
I guess the actual problem really is the Cisco router, but DropBox is still incredibly slow.
Re: (Score:2)
Re: (Score:2)
7:32p right now
Tracing route to ec2.eu-central-1.amazonaws.com 54.239.54.28
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms pfsense.localdomain 192.168.1.1
2 1 ms 1 ms <1 ms xx.xx.xx.1
3 1 ms 1 ms 2 ms xxxx [xx
Re: (Score:2)
Re: (Score:2)
Fancy version of FTP (Score:1)
What will they think of next? Remote desktop? It was called X-windows in the 1980s.
Kids these days think they invented sex.
Re:Fancy version of FTP (Score:5, Funny)
It was called X-windows in the 1980s,
That's X Window System to you, bub. That lawn you're on? It's mine. Off.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Actually, it's spelled rsync
Open Source alternative to Bittorrent Sync (Score:4, Informative)
is https://ind.ie/pulse/ [ind.ie] (was SyncThing).
Am I missing the point? (Score:5, Insightful)
They copied some data across a local network. Then they compared it how long it took to transfer the same data to remote servers across their internet connection? 1.36 GB in 41 seconds is 33 MB/s, which is either extremely underwhelming for local network performance (I suspect a magnetic hard drive bottleneck), or extremely impressive for a fat internet pipe, neither having to do with the software in question.
Re: (Score:2)
They copied some data across a local network.
I don't think they did, or at least it's implied - though not very clearly - that they didn't. In any case, the internet connection was 1gb/s, which is practically LAN speed with their gigabit adapters.
The article's author did a test over the internet and also found that Bittorrent beat the others - but then, the others are probably designed to be more considerate to your internet connection and not clog up your tubes.
Re: (Score:3)
"The transfer process was much longer. Times were in the double digit minutes, and largely depended on what connections my friends had."
In other words, in real-world scenario using the Internet, Bittorrent's Sync was not any faster than the times posted for th
Re: (Score:2)
Wait, why didn't you include this section?
"Yet it’s worth noting that Google Drive, OneDrive, and Dropbox still performed worse. They were limited by the same download bandwidth, but the upload section of the process was notably much slower (many ISPs worldwide offer much slower upload speeds than download speeds)."
So VB's test still gives the prize to sync. It's a bit weird that they didn't publish any times, though.
Anyway, I'm not saying Sync is obviously better, but your quote misses context that's
Re: (Score:2)
Indeed.
To further muddy the waters, DropBox supports (under Windows, at least) what it calls "LAN sync," with the goal of having data traverse LAN-WAN only once, no matter how many LAN clients want that data.
I do not know if it is default behavior. But I've seen it work just fine.
Re: Am I missing the point? (Score:2)
You think that would be a standard feature, but apparently it bears special mention.
I miss the older Foldershare then Live Mesh for that very reason. I think it might have been before "cloud" was a buzzword, and folks still thought about networks and file storage in a traditional way.
Skydive came out and I was fine with the giveth, but then was the taketh away. I remember being excited about the Live Framework developer API. The ideas presented don't seem especially innovative at the end of 2014, but they
Re: (Score:2)
Dropbox, at least, can do LAN syncing between devices. Of course, I didn't read the article, so I have no clue if they had it enabled on the computers so that it could be used, but based on the results, I'd doubt they did.
Re: (Score:2)
33 MB/s is pretty good even for a LAN link. But I think for most users, this may be pretty much irrelevant because for most people, the bottleneck is between your modem and your ISP.
iCloud Drive? (Score:1)
So they're using Apple hardware, but never tested the "new and improved" iCloud Drive?
I guess that was probably not released when they started their test.... I bet it would be equally slow though (especially since I think that part of iCloud actually runs on Azure).
Cloud service bandwidth caps, rsync? (Score:1)
All this has done is catalog what the bandwidth caps for the various cloud services are. The article itself admits that. BitTorrent performance is completely irrelevant.
A relevant comparison would be against other peer-to-peer transfer utilities like scp and rsync (w/ and w/o -z).
btsync and large shares (Score:1)
I've been using btsync to share my plex data elsewhere for backups. It works well, but what I've noticed is it slows down a lot when syncing more and more data.
It started out at 60 or so megabytes a second from a 10TB volume to an empty directory on a system on the local lan. After it passed a few TB, it slowed down to 40 megabytes a second. The last couple of TB took a long while as the transfer was down to about 10 megabytes a second.
This is still acceptable, because my outbound internet connection isn'
Infinit (Score:1)
Got to love synthetic tests (Score:1)
Encrypted node (Score:3, Interesting)
I've been using BitTorrent Sync for a year or so now. The main feature that was missing for me was the ability to set up an untrusted node which does not get access to the unencrypted data but can serve as a fast 24/7 proxy and backup system.
This functionality has now been added, although it's still in beta and only officially available in the API, not in the client... but a very simple hack makes it available in the client. This opens BitTorrent Sync open to 3rd party sync providers or cheap VPS.
The interface is still a bit quirky and designed for techies, but has also improved over time. Overall very happy with BitTorrent Sync.
No notification of concurrent modification (Score:3)
I have been using bittorrent sync for about the same amount of time, and the thing that is killing me is that it makes no effort to detect and warn when a file has been modified on multiple computer since the last sync. It just chooses the one that was modified most recently, and silently overwrites the other one. It does create a temporary archive backup of the modified file that was overwritten, but by the time you noticed you have lost data, it can be very difficult to wade through all the archive files
How secure is that connection string? (Score:1)
BTSync uses a string to connect computers and then sync files. How safe is that? I could start a script that tries to find these strings. If I find them, I can sync all files just like that. While that string is probably more unique than a username / password combination used on Dropbox, I guess Dropbox will see it when there are many failed tries on one username, or many failed tries from one IP. If your just guessing username and password, you can of course change usernames continuously, to avoid testing
Re: (Score:2)
Even if BTSync were to process one connection string per CPU clock cycle, it would still take 1e20 years to try all the possible 20-character Base64 strings that BTSync uses by default. If you choose a longer string, then it will take even more time. In otherwords, the standard strings have 120 bits of entropy, and you can increase that to up to 240 bits. This is less than is typically used for encryption these days, but btsync doesn't have to deal with offline attacks.
Rather than key size, I would be more
Re: (Score:2)
Although, now that I think about it, I'm not sure that couldn't be spoofed if you knew a little bit about the other person.
Re: (Score:2)
I don't understand why everybody is like, "err, btsync sucks until you show us the source!" The statements about closed source are true, but all the other major options are closed source as well so it's not a good thing to criticize btsync on!
I tried it out, and I liked it very much. I hit two problems, one of which is easier to solve:
1) no versioning implemented. The client could get really smart about versioning in a later update. This is primarily why I use dropbox, even on a machine where the files stay
Re: (Score:2)
Re: (Score:2)
We're a small company with just one overworked IT guy, so I don't think he was actively checking the logs and monitoring which computers were trying to access torrent ports. I chose not to draw attention to the fact that I was trying to torrent files (regardless of the justification), let alone be a person who was seen as complaining about it. That would raise eyebrows and be sent up the flagpole.
I ended up using Sync.com, which is a dropbox competitor based in Canada. I'm not naïve enough to think tha