can cofirm sky190176 is right.
Waittime 100 does work fine.
If I use 75 or 125, I get duplicate users and all reshares don't work anymore, but with waittime 100 everything works fine.
Maybe someone can explain howto finetune the waittime setting ??
can cofirm sky190176 is right.
Waittime 100 does work fine.
If I use 75 or 125, I get duplicate users and all reshares don't work anymore, but with waittime 100 everything works fine.
Maybe someone can explain howto finetune the waittime setting ??
delete post
It's useful for zapping users, as the ECM's are available in the cache directly the reader isn't queried. And zapping is faster that way, eg. Canal Digitaal ECM time is around 300-400ms, from the cache it's a matter of 10-20 ms.
Other than that: it's useful to prevent freezes on NDS, a little late ECM from another peer can be compensated with a cached ECM which comes in earlier (and hopeful just in time).
delete post
Ok will build and have a look now. Thanks
So far so good, no mem leak and cache is running, but still not just reading cache only seams to read card as well, however this is less I think. Will monitor for rest of the day.Originally Posted by Qnapper
Is csp settings still needed for the latest builds?
waittime, cachetime and cachesize have now become global settings.., so yes.. still needed.
needed?
csp settings used to be needed even if you didn't run a csp server so you could set the wait_time ... since we got those settings in the global cache settings now, you don't necessarily need to set the csp settings anymore ...
i think that's what he was asking for :)
I think i have bad experience with cache exchange!
Why? Because CW DIFF are a probblem unexplained so good do make me understood the process!
let's make thing clear and clean for everybody!
The ammounth of cache-ex don't matter for me, but only that can i use from my peers!
Someone told me i don't undersoot the mechanism of cache ex!
Fairly i want someone to explan how does it work!
Someone told me that when i doubt his change:
"i pushed over 4000 cache entries to you, your stats only show 50 received ... u do understand that means 3950 cache entries got ignored because you already had them in your cache? (and no, ignored cache entries are not counted in the received "got" stats).
it means those 50 cache entries were cache you did NOT have ... whether it resulted in a hit or not is completely irrelevant?"
I'm non soddisfied with his answer!
under i put this thing from 3 peers :
REVERSE CACHE PUSH 0 52369 28 0 141
REVERSE CACHE PUSH 0 1229 0 0 61
REVERSE CACHE PUSH 0 15689 184 0 15
Wich are good and wich are bad!
First collumn: amounth of cache recived( pull) second HITS, last CWdiff!
Please some explanation and tests!
In my oppinion:
First peer send me non only filtered caid( on this side he has not filtered anything) or send me 3-or 4 hops when i say clearlyr only hop2 i accept!
he sent me good dw because i have hit from his cache!
Second peer are a bad one because he send me nothing good and delayed CW(cw diff)This reader are candidate to be deleted!
The third one are good for me!
Has avverage, good hit and less cw diff!
CW diff are results of bad settings on the peer side, using a csp/multics or the last version of oscam compiled with multics instance!
I have blocked on the both side user/reader the csp ( drop csp) but how i get ride from this things?
How can i decrease cw diff?
Posting Permissions
Reply With Quote