kos13

Can you please tell me more about use case and why it is important to you? I see too many issues with transferring file ownership in p2p environment. I'll give you few examples: - Linux machine doesn't have the same user in system, what Sync should do? If you will add the 3-rd Linux box, what should be uid for the file? - You have the uid but don't have permission to change it. - You changed uid on one Linux machine, should it be propagated to other? I understand how this should work if you own and manage two Linux boxes. In p2p environment, different OSes and different ownership of machines it is unclear to me how Sync should behave. I completely agree that for enterprise world this is must have feature, and we will do it.

Done. Thanks.

Could you please follow these steps and send us logs. We are going to try this in test lab to see if iTunes does something fancy.

No. Right now it won't be configurable.

This is a bug. Will be fixed in next version.

Yes, if you have at least one side with port forwarding, Sync will establish direct connection.

Logs are always useful, this way we at least could verify that this is the same issue. So it is always better to have logs or any other clue that might help us.

We already received report, so the question was more to community, I believe. Problem is that if you enter Read Only secret for directory that contains data, this data will be added and will be distributed across peers. We know this bug and it will be fixed. One more comment. Sync is still in Alpha and we do work hard to fix the issues.

Could you please follow these steps: ? Looks like a bug.

There are few things: - You don't always know what external port is assigned, so when you report to DHT it will use your internal port; - DHT is slower than tracker, and we saw some models that drop the mapping - Tracker does help us to detect NAT type, so in future we could increase % of direct connections; - DHT traffic is not encrypted, so some companies (ISP) may block it. So using DHT is close to tracker, but tracker will give you better chances for direct connection.

@plop1602: Works on Synology DS 212j We don't have any specific steps for this NAS, please try to contact @plop1602 for details Thanks

We got report from user, that he was able to start Sync as service on Windows Server 2008 R2 using following instructions. http://support.microsoft.com/kb/137890 This is not generic case, but might help you. Instructions: 1. Copy instsrv.exe & srvany.exe to C:\Program Files (x86)\BitTorrent Sync, you should get these files from Windows NT Resource Kit 2. In Command Line from Administrator execute following: cd C:\Program Files (x86)\BitTorrent Sync instsrv BTSync “C:\Program Files (x86)\BitTorrent Sync\srvany.exe” regedit 3. In regestry editor find HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\BTSync Create key: Parameters leave key:blank Inside Parameters create new value (type REG_SZ): Application with value: C:\Program Files (x86)\BitTorrent Sync\BTSync.exe 4. In application preferences make sure to unchek "Start application when Windows starts" 5. Close BTSync 6. From the Services applet in Control Panel find BTSync make sure that type is Automatic and inside "Log On" tab enter your login and password. 7. Start BTSync service. Please let us know if this solution will work. Thanks to Ivan Nikitin. Original article in Russian http://ivannikitin.c...ittorrent-sync/

Idea is that if you want to go completely server less, DHT is one of the option to connect peers over internet. Pros: is that DHT is always accessible and is not hosted by any company. Cons: DHT is bad in doing hole punching, so some peers won't be able to connect directly.

We are looking to this case. Thank you for logs.

Sync right now can support up to a 1M files. However it will require significant memory for that. We understand how we could improve that but not there yet. Memory usage for 400k files might be in range of 1Gb = which is a lot. We are working on improvement. Please also note, that there is a virtual memory and real memory. Usually all memory tools shows you virtual memory.

TCP is well optimized for each OS. There are a lot of kernel optimized things that is happening inside kernel. While UDP is simpler from OS perspective, Sync has to do a lot of uTP activity at user level. This will involve context switching, data copy from kernel to OS etc. This will add overhead especially on low end CPU like NAS.

Sync uses upd for internet transfers. So most likely upd is blocked somewhere on your route. As it sees UPD is not blocked by Great FW of China, it must be somewhere inside your network. You may also try to follow these steps - and we will give you better clue.

From now on.

We will fix that

Sure. We will add this.

Thank you. We will need debug logs to find issue. This is odd.

And if this is implemented, what will be a use case for such functionality?

Yes, Sync re-scan folder every 10 minutes. There is no option to turn this off.

Please, it would be very interesting. We never thought about such use case. API is so generic terms that different people might use in different context. Therefore we asked you what you want to see. This will give us better chance to develop something really usable.

We will check that. The only Sync is doing is rescan directory every 10 minutes, and then write new statys to some file. I am not sure how Windows will show that. We will check and get back on this.