Opened 12 years ago
Closed 11 years ago
#2229 closed enhancement (wontfix)
Users can use the cache of a reader with different group
Reported by: | Kayvan | Owned by: | corsair |
---|---|---|---|
Priority: | major | Component: | General |
Severity: | medium | Keywords: | |
Cc: | Sensitive: | no |
Description
Revision
6059
Issue Description
Under the circumstances, users can use the cache of a reader with different group
When the issue occurs
If a group is shared between them
Attachments (3)
Change History (10)
by , 12 years ago
by , 12 years ago
Attachment: | oscam.server added |
---|
by , 12 years ago
Attachment: | oscam.user added |
---|
comment:1 by , 12 years ago
Owner: | set to |
---|
comment:2 by , 12 years ago
comment:3 by , 12 years ago
It wasn't like this before, this will cause a problem when USR2 sends the ECM earlier than USR1 to the reader ( Only group 32 ) and it can't decode it, then USR1 get the same result, Because it uses the same cache and none of the ECM(s) would sent to the fallback reader(s).
comment:4 by , 12 years ago
Resolution: | → invalid |
---|---|
Status: | new → closed |
oscam has one global cache. each user on the server can access the cache and it does not matter which group he is in. this is since a very long time like this.
if user one sends a request to the server he is not allowed the request is denied by a filter (group, caid, etc.) and won´t generate a not found. if user two requests the same channel and he is allowed to request it normal porocessing takes place.
comment:5 by , 12 years ago
Resolution: | invalid |
---|---|
Status: | closed → reopened |
As written here http://www.oscam.cc/wbb3/index.php?page=Thread&threadID=17334 i think the ticket is valid
Say you have two users in different groups:
- USR1 (Group 1) has access only to proxies
- USR2 (Group 2) has access to the local card
When they are both watching the same channel, following can happen:
2012/09/10 11:44:09 1020600 c USR2 (XXXX&000000/0000/ZZZZ/58:450C): found (144 ms) by localcard (L/1/2/2) - ChannelXYZ 2012/09/10 11:44:16 1020600 c USR2 (XXXX&000000/0000/ZZZZ/58:3F39): found (131 ms) by localcard (L/1/2/2) - ChannelXYZ 2012/09/10 11:44:23 1020600 c USR2 (XXXX&000000/0000/ZZZZ/58:5E37): found (129 ms) by localcard (L/1/2/2) - ChannelXYZ 2012/09/10 11:44:33 101C200 c USR1 (XXXX&000000/0000/ZZZZ/58:3A96): found (610 ms) by PROXY - ChannelXYZ 2012/09/10 11:44:33 1020600 c USR2 (XXXX&000000/0000/ZZZZ/58:3A96): cache2 (614 ms) by PROXY (F/0/2/2) - ChannelXYZ 2012/09/10 11:44:47 101C200 c USR1 (XXXX&000000/0000/ZZZZ/58:42C6): timeout (10000 ms) by PROXY - ChannelXYZ 2012/09/10 11:44:47 1020600 c USR2 (XXXX&000000/0000/ZZZZ/58:42C6): timeout (9997 ms) (F/0/2/2) - ChannelXYZ 2012/09/10 11:44:57 101C200 c USR1 (XXXX&000000/0000/ZZZZ/58:0FE9): timeout (10000 ms) by PROXY - ChannelXYZ
USR2 gets the slower ECM-Times of USR1 and even the timeouts. This shouldn't happen, because USR2 has access to the local card.
Actually the cache should be shared, but USR2 shouldn't depend on the USR1 request.
Ideal would be:
- USR1 makes his ECM-Request via proxy
- USR2 makes his own request via local card
- USR2s Request is faster than USR1 an is written to the cache
- USR1 can use the faster answer via cache as long USR2 is on the same channel
I'm using Oscam SVN version r7362 on a MacMini running SnowLeopard.
comment:6 by , 11 years ago
Type: | defect → enhancement |
---|
Lets call it a feature, in stead of a defect ;-)
comment:7 by , 11 years ago
Resolution: | → wontfix |
---|---|
Status: | reopened → closed |
creating a own cache for each group would be a massive and in my opinion useless overhead. the only solution for your setup would be to create 2 separate oscam installations. one with the cards and one with the proxie readers. this way you can separate the traffic.
I think so what??? I do not think this can be blocked on the cache level...