similar to: Queue help: Extending RRMEMORY strategy to use penalty

Hi, I would like to implement a queue using a circular strategy, I mean, using roundrobin or rrmemory strategies. However, I am not able to define the order Asterisk will call the agents once a new call arrives in the queue. Seems that Asterisk will always define its order as the queues.conf file is read, and most of times this order is different from the one I want (for each queue in

Dear all, I can't understand the diff between roundrobin and rrmemory strategy. Could you explain for me ? and is roundrobin means each available interface ring once or several times and ring another? ; A strategy may be specified. Valid strategies include: ; ; ringall - ring all available channels until one answers (default) ; roundrobin - take turns ringing each available interface ;

Hi, I'm trying to use Round Robin Memory with my queues. It seems to work fine... that being I call in.. first time agent 1 will get a call, second time agent 2, and so on. However, it seems if a period of time has gone by since agent 1 got a call (I don't know how much but say 10 minutes or so) when another call comes in.. agent 1 gets the call again. Can anyone confirm this? Is

Hello, I've been using * for quite a while using AVM C4 card and a few Grandstream Budgetone 102 hard phones plus couple of HandyTone 286. Echo is a big problem. I am getting used to it, but some users complain. Anyone has experience w/ regards to echo comparing the AVM C4 with two HFC-cards? Before shelling out time and money, maybe someone else already has done so, and could tell me if

I am using rrmemory for my queues. I have noticed that the application will only distribute or dial one number at a time. Is there a different strategy that will allow the queue to distribute more than one call at a time? I don't want to use ringall because that would tie up thirteen of my trunks every time it tried to distribute a call. Any thoughts? -------------- next part -------------- An

I have set up an operator queue for our receptionist. That way, if she takes a break or is out, by logging out of the queue, calls to the "Operator" can be handled by other agents. I have set strategy = roundrobin in queues.conf. According to "the book" ATFoT, roundrobin always starts with the first agent in the queue. This is the desired result. I want all calls to start

Hi, I have configured a call queue as follows: [sales] music = default strategy = roundrobin timeout = 20 retry = 5 reportholdtime = yes announce = queue-sales member => Agent/9021 member => Agent/5901 member => Agent/9020 member => Agent/1114 Now, I would expect all calls to this queue to be delivered first to 9021, then to 5901, then to 9020, then to 1114 (all agents are logged in

Hi, I am using Asterisk 1.4.18. I am implementing Penalties for my agents. What is happening: two agents configuired one agent with penalty 1 and the other with penalty 2. All the calls must go first to Agent 1 and if his line is busy then only then agent 2 will get the call. However my queues are not behaving in this manner. I have impmemnted ringall strategy. Now when first call comes it

Hi all, When implementing a queue using members like this: [queue] strategy => rrmemory member => Agent/1000,1 member => Agent/1001,1 member => Agent/1002,2 member => Agent/1003,2 member => Agent/1004,2 And you call into the queue, agents 1000 and 1001 will ring in an alternating fashion until one of them answers it. You might have seen my question coming already, so I

Howdy, have working realtime queues using queue_members looking something like; queuea|Local/101@context|0 queuea|Local/102@context|1 queuea|Local/103@context|10 Regardless of what strategy is used in the queues (roundrobin,rrmemory,ringall etc) it wont escalate on NOANSWER Asterisk SVN-branch-1.2-r33841 Any clues are appreciated! /Danny

Seeing as the "leastrecent" strategy does not work for us (it will *always* call the leastrecent agent, even if that agent is busy, and will not move to the next agent) I thought I'd try the roundrobin strategy. This seemed to work - however, I had my supervisor telling me today that a certain agent seemed to be getting most of the calls - and when I checked, out of 50 inbound

Just woundering if the intentend functionality of leastrecent and fewestcalls it to continually dial only the first chosen ext. in the queue. In other words if a memeber is logged into the queue but doesn't answer the call the call never moves on in my configuration from that ext. This could be really bad!!!! Thanks [support] announce-frequency=45 strategy=leastrecent music=default

I attempted setting up a queue with agents that log in, and get called with incoming calls: Agents log in using: exten => *88,1,AgentCallbackLogin(${CALLERIDNUM}|${CALLERIDNUM}@test-sip) Calls get into the queue with: exten => 6029995654,1,Queue(test-noc|t|||60) queues.conf: [test-noc] strategy = rrmemory context = test-sip timeout = 10 retry = 4 member => Agent/@2

Hi there, this is my queues.conf, I'm using todays CVS: [599] joinempty = yes musiconhold = default strategy = roundrobin servicelevel = 60 wrapuptime = 0 maxlen = 0 timeout=15 announce-frequency = 15 member => SIP/381 member => SIP/300 At first call 381 rings, if you hang up and call again you get the 300 ringing ... this looks more rrmemory than roundrobin, there is something wrong

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 We're running multiple instances on the same server. While doveadm penalty works: root at host:/etc/dovecot# doveadm penalty IP penalty last_penalty last_update 93.180.xxx.xxx 4 2014-05-10 12:21:08 12:21:08 80.153.xxx.xxx 0 2014-05-10 12:21:12 12:21:12 a seperate anvil-Socket doesn't work at all: root

Since the changesets: http://hg.dovecot.org/dovecot-2.0/rev/9bec70f26637 http://hg.dovecot.org/dovecot-2.0/rev/75b48fa3aec4 there is another problem. A `dovecot stop`, while auth and auth -w are still running, logs: Nov 12 03:17:00 master: Warning: Killed with signal 15 (by pid=31656 uid=0 code=kill) Nov 12 03:17:00 auth: Error: read(anvil-auth-penalty) failed: EOF Nov 12 03:17:00 auth:

Hi Timo, on a 32bit Debian squeeze/sid system, with gcc-4.4 (Debian 4.4.2-9) 4.4.3 20100108 (prerelease) I'm getting this warnings: penalty.c: In function ?penalty_timeout?: penalty.c:169: warning: comparison between signed and unsigned integer expressions test-penalty.c: In function ?test_penalty_checksum?: test-penalty.c:27: warning: comparison between signed and unsigned integer

Using doveadm penalty I just noticed a penalty for 0.0.0.0. Is this Postfix querying the auth-socket for his SASL-requests? I tried adding 0.0.0.0 to login_trusted_networks, but that doesn't help. Peer -- Heinlein Support GmbH Schwedter Str. 8/9b, 10119 Berlin http://www.heinlein-support.de Tel: 030 / 405051-42 Fax: 030 / 405051-19 Zwangsangaben lt. ?35a GmbHG: HRB 93818 B /

Anyone have experience specifying the "penalty.factor" option in the "glmnet" command? I have 3 variables (out of a million genotype variables) that I want to force into the model (i.e., set penalty factor to 0), but I can't figure out how to do that. [[alternative HTML version deleted]]

It's so you can have agents that are less likely to take calls (e.g. imagine a sales queue, you'd have the sales people with no penalty, you might have the receptionists with a penalty of 1 and us propeller heads in technical support with a penalty of 2). The technical support people would only be offered a call from the sales queue if all the sales people and the receptionists were busy.