11 C Lines - 1 dedicated Debian running OScam - 4 Dreamboxes - Best Setup?

There are 9 replies in this Thread which was already clicked 3,001 times. The last Post () by damnedfreak.

  • Hello all,


    First, I will explain my setup.


    I have 11 C Lines from friends. To be honest, I am not sure they have local cards. I'm afraid they don't.
    I have 4 Dreamboxes, all in my network.
    I have 1 dedicated Debian running OScam, located in a real data center. Before, I was running a dedicated server in my network, but I found the ECM times are better with the data center server. Probably due to the data center having much better connectivity than my standard line.


    Initially, I had 4 C Lines, each Dreambox having one line setup. It worked okay, but with some freezes, especially on the weekends when there are football games in Europe (Germany, UK, ..)
    My plan was to combine all the C lines in a server and distribute them to the boxes. My plan was that even if one line has a freeze, the others will still serve the ECMs and the boxes connected to the server will not notice anything and work seamlessly.


    So I added the 4 Clines to the dedicated Oscam server and added 7 more lines I gathered from more friends. They are willing to give me their C Line so that they can connect back to my server, taking advantage of the combined lines.


    So now I have 11 C Lines going to 1 dedicated Oscam server, which serves 4 CCCam clients on Dreamboxes.
    I have less freezes on the Dreamboxes, but still some. Which I find a bit frustrating, since I am prety sure that not all 11 lines fail at the same moment. So it is most probably my config which needs some fine tuning.


    My oscam.conf is as follows:



    My oscam.servers:



    My oscam.server (just putting 1 line, it is 11 times the same just with different servers and users):


    My oscam.services:



    The Dreamboxes run CCCam 2.2.1 and are connected as follows:


    Code
    C: XXXXXXXXXX 31337 user1 user1
    C: XXXXXXXXXXXX 31337 user1 user1
    CAID PRIO FILE : /var/etc/CCcam.prio
    PROVIDERINFO FILE : /var/etc/CCcam.providers
    CHANNELINFO FILE : /var/etc/CCcam.channelinfo


    My goal is to eliminate as much freezes as possible by always delivering the ECM from any of the 11 C lines to the receiver.
    My ecm times varying between 100-1000 ms.


    I see there is a debug mode in Oscam, is there a good tutorial on how I can use it to debug my setup? Or should I run some kind of debug on the CCCam clients?


    Hope to get some constructive advice from you guys!



    Regards,
    Steve

  • I think that not abusing your peers trust by ignoring reshare levels would be a good start!


    I understand your point. Thank you for your ethical advice. If it helps you, just blend out the cccignorereshare line in my config.


    However, I am still looking for technical advice.
    I have an example of the oscam log, where one of the clients experienced a freeze:



    user1 experienced a timeout on reader2 (1600 ms, which is what I have setup in the global conf of oscam) - so far so good. But now, it is getting a bunch of timeouts by reader3, and then a cache of reader 1, and only 6 seconds later a found by reader1. Why is it trying the reader3 a couple of times and then only the reader1 after 6 seconds? Is there a way to set it up to lookup additional readers and faster, so that the freeze is eliminated or at least highly reduced?


    Thanks to those in advance who are willing to answer my technical question despite the bad start of the thread. The question is a valid one anyways, I believe.


    Regards,
    Steve

  • I dare to say I get that. However, if one of the cards I receive is freezing (for whatever reason: because it's reshared too much, because of God's will, whatever?), Oscam should not try to access it a dozen time but switch to another one which is not frozen. That's my understanding of load balancing or better fallback.


    As you can see in the log above, one card fails, then Oscam keeps querying another card which also seems frozen. Why is it not going after another card, since I have more available?


    I hope you don't get me wrong, but that should be true regardless of the origin of the cards :)

  • Thanks, will look into that.


    What about the lb_nfb_readers parameter?
    I had it set to 1 (I think by default).
    So for each active CAID, does it means it has one reader which will serve the ECM, and one reader ready to serve same ECM if the primary reader failed? That would explain the case above, where reader2 timed out, reader7 (which I presume was the only fallback reader since the paramter is set to 1) was also not able to serve the ECM, which in turn resulted in the freeze until the next ECM was successfully delivered by reader3.


    If my assumption below is true, increasing the value of lb_nfb_readers will help in this kind of situation.


    Thanks for staying in the conversation anyways!


    Cheers,
    Steve

  • Why use oscam when you have no card of your own, other than stretching reshare there is no benefit really, have you tried CCcam?
    It would make things alot simpler for you imo.


    I suppose everybody has their own opinion and for different reasons.
    I use Oscam only for a couple reasons
    Ignorereshare . I believe that Ignore reshare should not be a global setting in configuration and should be used only on individual user basis. I use a PC server and when I ask for a share I always make it clear. Most shares I get don’t have a reshare and in CCcam are useless to me. I argue that if someone gives me a share he/she intends that I use it so I feel it is legitimate for me to pass that share to my STB. Hence I think it is acceptable to use ignorerehare for my userSTB. I would never reshare to a peer where I did not have that entitlement and that feature should be changed in Oscam. It is a bad bad reflection on cardsharing.
    Stability. Having used Cccam for 3 years I found it very stable for sometimes up to 2 months. Then for no apparent reason it could’nt be run. Tried everything including contacting all peers and it couldn’t be fixed. Switch off for 24-48 hours and its back working normal without any change. So when all roumours about backdoor I started to look elsewhere.
    I have found Oscam difficult to learn but I am now using Camd35 with oscam and(apart from the pcsc) I find it extraordinarly stable and very well supported.
    Unfortunately I am not able to help with the original problem so sorry to damnedfreak for this little rant. It is a wet cold Sunday morning

  • I suppose everybody has their own opinion and for different reasons.
    I use Oscam only for a couple reasons
    Ignorereshare . I believe that Ignore reshare should not be a global setting in configuration and should be used only on individual user basis. I use a PC server and when I ask for a share I always make it clear. Most shares I get don’t have a reshare and in CCcam are useless to me. I argue that if someone gives me a share he/she intends that I use it so I feel it is legitimate for me to pass that share to my STB. Hence I think it is acceptable to use ignorerehare for my userSTB. I would never reshare to a peer where I did not have that entitlement and that feature should be changed in Oscam. It is a bad bad reflection on cardsharing.
    Stability. Having used Cccam for 3 years I found it very stable for sometimes up to 2 months. Then for no apparent reason it could’nt be run. Tried everything including contacting all peers and it couldn’t be fixed. Switch off for 24-48 hours and its back working normal without any change. So when all roumours about backdoor I started to look elsewhere.
    I have found Oscam difficult to learn but I am now using Camd35 with oscam and(apart from the pcsc) I find it extraordinarly stable and very well supported.
    Unfortunately I am not able to help with the original problem so sorry to damnedfreak for this little rant. It is a wet cold Sunday morning


    No need to feel sorry, actually I find your post really interesting ;-)
    I am using Oscam to "stretch" my clines for the same reason as you do. I have a server and connect all my boxes (not other people's!) to it, to take advantage of the loadbalancer. CCCam will not do that since my boxes would not get the shares, only the server. Where they are useless since I don't watch TV on my Debian in the data center :D


    And by the way, changing the lb_nfb_readers value to 4 greatly reduced the frequency and length of the freezes. Which is quiet obvious, since you can have 1000 lines on the Oscam server, when you leave that value on the default, the Oscam server will only attempt to send the ECMs from 2 different lines to your box for any requests. So if both fail or are too slow, you have your annoying "black moment". I am actually surprised nobody came up with that solution to my question.


    Anyway, maybe it will help other people in the future.


    Regards,
    Steve

OSCam Support Forum

Configs, discussion, downloads and guides for OSCam Softcam.

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!