Cacheex 3 (r 8687) stops pushing to certain clients

[gimic]

Hi,

i'm on release 8687 and i've noticed yesterday that the push towards some peers stops. i've got 8 cacheex peers, this night everything was fine. this morning only 3 peers were getting pushed cache. on the others the counter stayed the same. After a restart of oscam, the push begins again for all readers.

Does anybody have this? and if so, what is causing the problem and how can i help to solve this?

Tnx,
Gimic

[gimic]

Again, i needed to reboot...

this is what i find in normal log:

2013/06/08 10:03:26 1B5B1A0 c WARNING lock csp_lastnodes destroy timed out.
2013/06/08 12:17:31 0 --- Skipped 1 duplicated log lines ---

this is when i try to stop a reader, web interface hangs, icon changes that reader is stopped, but i hasn't stopped.

2013/06/08 13:25:24 D00008B0 h WARNING lock http_lock (WRITELOCK) timed out.
2013/06/08 13:28:03 D00008B0 h WARNING lock system_lock (WRITELOCK) timed out.
2013/06/08 13:29:24 D00008B0 h WARNING lock http_lock (WRITELOCK) timed out.

if nothing works, i do restart of oscam.
2013/06/08 13:29:29 D00008B0 h Restart requested by WebIF from 178.118.55.36

there is a ticket already for this problem posted on streamboard (http://www.streamboard.tv/oscam/ticket/3347) but don't know how to contribute.

Greetz,
Gimic

[gimic]

Could this be a problem?
the readers that have this error stop working. No explaination why. now it occured 5 min after reboot, they only pushed 5000 ecm's...

2013/06/08 13:44:23 CC0382D0 c add client job action 28 queue length 1 xxxxxx
2013/06/08 13:44:23 CC0382D0 c WARNING: job queue reader cacheuser1 has more than 2000 jobs! count=2001, dropped!
2013/06/08 13:44:23 CC0382D0 c WARNING: job queue reader cacheuser2 has more than 2000 jobs! count=2001, dropped!
2013/06/08 13:44:23 CC0382D0 c WARNING: job queue reader cacheuser3 has more than 2000 jobs! count=2001, dropped!

why does it not run jobs and fails somewhere?
this log came from activating all loggs, exept cachex log.

[turbopower]

Provide your configs attached.

[gimic]

This is my config,

and the config of the reader that is now not getting pushed ecm's and i get the message that there are 2000 jobs running

Tnx for looking into this.




Oscam.conf

# oscam.conf generated automatically by Streamboard OSCAM 1.20-unstable_svn SVN r8687
# Read more: http://www.streamboard.tv/svn/oscam/trunk/Distribution/doc/txt/oscam.conf.txt

[global]
logfile                       = /var/log/oscam/oscam.log
clienttimeout                 = 10000
nice                          = -1
maxlogsize                    = 30000
waitforcards_extra_delay      = 3000
preferlocalcards              = 1
dropdups                      = 1
usrfile                       = /var/log/oscam/usr.log
mailfile                      = /var/log/oscam/mail.log
lb_mode                       = 1
lb_save                       = 500
lb_retrylimit                 = 1600
lb_savepath                   = /usr/local/stat

[cache]
max_time                      = 25
max_count                     = 20000
wait_time                     = 10
csp_port                      = 12464
csp_allow_request             = 0
cwcycle_check_enable          = 1
cwcycle_check_caid            = 0100,17,18
cwcycle_maxlist               = 2500
cwcycle_keeptime              = 15
cwcycle_onbad                 = 1
cwcycle_dropold               = 1
cwcycle_sensitive             = 4

[reader]
label                         = cachereader
description                   = Satfriends 
protocol                      = cccam
device                        = xxxxx.xxxx.com,13000
user                          = xxxx
password                      = xxxxx
inactivitytimeout             = 30
cacheex                       = 3
cacheex_maxhop                = 3
cacheex_ecm_filter            = 0100
cacheex_drop_csp              = 1
group                         = 20
cccversion                    = 2.3.0
cccmaxhops                    = 0
ccckeepalive                  = 1

[turbopower]

[global]

logfile                       = /var/log/oscam/oscam.log
clienttimeout                 = 10000
nice                          = -1
maxlogsize                    = 30000
waitforcards_extra_delay      = 3000
preferlocalcards              = 1
dropdups                      = 1
usrfile                       = /var/log/oscam/usr.log
mailfile                      = /var/log/oscam/mail.log

[cache]
delay                         = 1
max_time                      = 13
max_count                     = 10000
wait_time                     = 250
csp_allow_request             = 0

Please use this config.

[gimic]

Tnx,
applyed the config like you suggested, left out the loadbalancer and the cwcylcle check. i'l keep you posted if it runs well now.

[gimic]

it keeps running now, so the settings above did the job, if it stops again, i'l post it here.

Tnx for your help

[bvgelder]

On which caid's is this based?

max_time=13
wait_time=250

[turbopower]

On which caid's is this based?

max_time=13
wait_time=250

About seca 0100 you may to increase this time because seca starting to freeze after 1000 msec but this will increase time on answers on your local or virtual cards.