Opened 10 years ago

Closed 9 years ago

#3471 closed defect (fixed)

Offline / Inactive reader still being queried?

Reported by: marrik Owned by:
Priority: critical Component: Protocol - CCCam
Severity: high Keywords: cccam ecm nok query offline reader
Cc: Sensitive: no

Description

Revision

r8959

Issue Description

When a reader goes offline it's still being queried with ecm requests

When the issue occurs

Always

How the issue is reproducable

Configure a reader (cccam) with wrog information so that it will never connect

Query a reader that isn't online results in ECM NOK statistics litter

Example reader config used:

[reader]
label = server
protocol = cccam
device = servername,port
user = xxxx
password = xxxx
inactivitytimeout = 30
reconnecttimeout = 30
caid = 0604
group = 21,22,23
cccversion = 2.1.3

Attachments (1)

Oscam_Offline_reader_queried.png (8.3 KB ) - added by marrik 10 years ago.
Reader offline being queried

Download all attachments as: .zip

Change History (9)

by marrik, 10 years ago

Reader offline being queried

comment:1 by stefansat sat, 10 years ago

Priority: minormajor
Severity: lowhigh

I can confirm this, and it is not only offline readers, when rate limit is actif on a card
also there is an loop starting to try and retry eternally. That's processor and log files consuming . And it's a bug. Do not now anymore since with version this issue started.

Setting priority to high.

comment:2 by theparasol, 10 years ago

Do my ecmhandler patches fix this? (See forum)

comment:3 by marrik, 10 years ago

will test the ecmhandler fix soon.. But just noticed it's only happening (at least on my side) for the reader with the id -> 2.

Edit: it's not only reader with id 2... (it happens to all readers.. ) :-)

Last edited 10 years ago by marrik (previous) (diff)

comment:4 by marrik, 10 years ago

Which ecm handler patch exactly?.. (Searched but can't find it.. )

comment:5 by stefansat sat, 10 years ago

No I do not think ecm handler patch will solve this. It's eighter started after blue eleven patches. Also it's maybe the cause off memory leak described on ticket 3441

As if a reader is set to off when a client made actif use of it an eternal loop is started with ecm not found.

Also when a card is pulled out same happens.

Also when rate limit activates same happens.

It is very very agresiff and fill ups logs into an almost unreadable log file (as it becomes so big) within a single minute.

I'm shure it's somwhere a single wrongly set accolade or an accidentally erased && statement. But to find this , That's another problem.

comment:6 by stefansat sat, 10 years ago

Priority: majorcritical

Some extra

It has nothing to do with dvbapi.

Problem is really server based and happens to all clients.

Problems occurs for local cards into, smargo reader (smargo or smartreader protocol), local internal readers (dm8000), pcsc readers ubuntu pc 12.04.

Whit other words simply I dear to say all card readers. All clients and server types off oscam.

I did my best but could not trace from when it started.

However this toppic seems to me very very important as somewhere a very severe bug is in.

I increase this toppic to to highest error level possible here it's a really almost fatal bug.

comment:7 by marrik, 9 years ago

Follow up: At this moment build r10122 doesn't seem to have the issue anymore. If someone can confirm this? The ticket can be closed?

comment:8 by stefansat sat, 9 years ago

Resolution: fixed
Status: newclosed

As far I can see it does not happen anymore. I'll close the ticket.

Note: See TracTickets for help on using tickets.