Beiträge von joaoabs

  • How safe can OMV be with Owncloud ports accessible from Internet

    Hello,


    I successfully configured OwnCloud in my OMV.
    When I'm at home, my smartphone pictures are automatically uploaded to the NAS.


    Now, I want to be able to do the same whenever I'm out. I know that I would need to forward some ports in my router/firewall to the OMV machine. My question is, even if I use the shttp port (443) how safe would that be? Would debian/OMV/Owncloud be strong enough to avoid attacks?


    The point here is, even if the password is strong and even if using a certificate to connect, the fact that a port is opened inside my LAN could be a security flaw, and could be exploited. I've heard some stories that the simple fact that an SSH sever was secured (higher port, login with certificate, no root login, etc.) but reachable from the internet was enough to get hacked.


    What is your opinion/experience?


    Thanks,
    Joao

  • kworker process with high CPU usage

    Hello,


    I'm using OMV for one week or so, and I'm still migrating my data from my old NAS systems that weren't not performing as expected any more.
    I'm using 2 main processes in parallel to migrate the data:
    - Using MidnightCommander in OMV to mount the remote NAS via FTP and then copping it;
    - Using my Ubuntu PC that has SMB mounted to all NAS and basically dragging and dropping (I know here that the data comes firstly to my PC and then to OMV, theoretically affecting performance, but strangely it is as fast as direct FTP.).


    In neither of these processes I'm able to achieve decent GB speeds (according to system information graphs, the max peak ever registered was around 260mbps), although my home network is cat5e gigabit. Anyway, I suspect the bottleneck is in the old NAS systems, so I'm not blaming OMV :).


    So, this migration process is taking its time, some batches of copies take couple of hours. When this happens, I notice that there is a process in OMV that starts consuming lots of CPU (around 100% of one CPU core) and remains like that even after the copies are finished. The only way to efficiently recover is rebooting the system.


    I investigated a little and this was a known bug in older kernel releases (saw bug reports on 3.0.0-12-generic, 3.13.0-32.57, etc.). It should be solved in the current release (OMV 2.1.27 --> Linux 3.2.0-4-686-pae).
    I couldn’t find any reference here in the OMV forum.


    So, having this in mind, I'd like to ask if someone else is also suffering from the same symptom, and if there is a solution for it.


    Thanks,
    Joaoabs


    ---------
    Build:
    headless JNF95A-270-LF Intel Atom PC (1st Generation Intel® Atom N270 @ 2.5W (512K Cache, 1.60 GHz, 533 MHz FSB) with 2GB RAM and GB ethernet)
    Data on Western Digital Red (NASware) 3TB
    OMV 2.1.27 running on SanDisk UtraFit 16GB with flashmemory 1.9.2 folder2ram plugin

  • Improve PLEX transcoding on a low power PC

    Hello,


    I'm a pretty recent user of OMV and so far I can say I'm still on the honeymoon. :)


    I'm running it on an low-power, headless JNF95A-270-LF Intel Atom PC (1st Generation Intel® Atom N270 @ 2.5W (512K Cache, 1.60 GHz, 533 MHz FSB) with 2GB RAM and GB ethernet) and in terms of NFS and CIF/SMB sharing it's working great: Much faster than my other NAS (WD-MyCloud).


    Anyway, I installed the OMV-extras PLEX plugin but unfortunately my system isn't powerfull enough to perform the transcoding (the graphic card is Intel GMA950 Integrated Graphics).
    However, I found some mini-pci-express video decoder boards for sale in ebay like this (http://www.ebay.com/itm/BCM700…2068a7:g:FLMAAOSwBahVFR~Q).


    Does anyone knows if this kind of boards can be used/compatible by/with OS/PLEX in order to improve the transcoding?


    If not, would a "normal PCI" graphics board (evethough my system is headless), help the transcoding? If so, would it work "automagically", or would it require any special configuration?


    Thanks,
    Joaoabs