i_q

Members
  • Posts

    3
  • Joined

  • Last visited

i_q's Achievements

New User

New User (1/3)

  1. Probably no real answer/help for you, but I know this behaviour from btsync when the system is 'a bit underpowered' and the shares it has to index contain _a_lot_ of files ... ... then the bootstrap of btsync seems to take a very long time, and in this time, the webinterface doesn't react at all (but the systems kernel already accepted the TCP-Connection. Hence no timeout) So often enough it seems simply enough to just wait until it has done its thing. When I switched to latest btsync 2 it also took me a very while to react. So I was interested to find out what it's doing and startet to watch the open filehandles with "lsof" What I found for _MY_ system, was, that I was syncing a folder, which contained some symlinks into "/" And btsync seemed to follow them recursively. The problem when I catched it doing this, was, that it seemed to get stuck somewhere in $MYSHARE/some_directory/symlink_to_root/proc/$SOMEPID/root/proc/$SOMEPIDso it turned nice circles indexing all the files in /proc without noticing .... Thats something I haven't seen before with 1.4.* Maybe you're experiencing something similar ...
  2. thanks GreatMarko, for the answer ... So Iĺl probably have to play a bit with this. I'll see, if I can work out a linux-solution for this ... ... currently I'm thinking about running multiple instances of btsync via lxc - anyone tried, what happens, if different btsync-instances are seeding the same directory?
  3. btsync is great - thanx ... But I have a small issue I have some boxes in the datacenter, where BTsync is running and seeding to multiple hosts. normally I'll put new stuff on one host and these will be seeding to all the clients. So now I have these hosts connected with 1G links to the Internet and I have connected them on the backside with a dedicated 10G Network (Yes, the machines have 10G links) So my Idea is, that I push some new files to one of the seeder, they will use all the bandwidth between them and then seed it to "the world" with an 1G uplink each. But as it seems, the 10G network won't be used. Even if I specify the IP-Adresses on the second Interface they don't seem to find each other on this link. The sync rate never goes >100MB/s per host, and OMG this hardware _WILL_ do a lot more. As this is currently an prototype and I'm planning on using this for intercontinental sync between datacenter-locations it's important for me, to test if this works - This setup can also be replicated in a small environment, if you plug two workstations two an WIFI-Network for sharing/seeding, and connect them on the backside with a straight Cat5E-Cable. Configuring some IP-Adresses on the secondary net, and giving them as an seeder will not use the cable link but rather still use the wifi. Of course, if you're gathering enough hosts/clients on the net, that are reseeding the stuff, it all will be fine, but the bootup-phase is critical and can possibly be much more accelerated this way ... Any ideas?