vicidial.org

i'd like to know what other ways i can improve call quality and reduce wastage. currently, i have:

goauto quad core2, 2gb ram, 250 basic hdd, and 2Mbps leased line.

recordings are performed in eyebeam and configured to use g729.

if i move to auto record, will my performance degrade significantly? can my setup accommodate 30 agents on manual dial, recording done in dialer?

anyone?

The specs you mentioned can do full recordings, g729 codec, 3:1 dial ratio with 30 agents smoothly. You just need to do some optimizations. Enable RAMdrive or tmpfs. Mix the recordings in MP3 during off hours and etc.

when you post, please post your entire configuration including (but not limited to) your vicidial version with build.

Please do not say "standard install". Remember that you are asking for help from a group of people who may or may now know what YOUR idea of a "standard install" is, and yet may have LOTS of information helpful to you IF you post your specs.

Try something like this:

Similar to This:
GoAutoDial Version X.X from .iso | Vicidial X.X.X-XXX Build XXXXXX-XXXX | Asterisk X.X.X | Single Server | No Digium/Sangoma Hardware | No Extra Software After Installation
________

If you are speaking of MANUAL dialing, 30 agents should be good. If you want AUTODIAL 3:1 ... you may need another server in the cluster.

Do enable RAMdrives (perhaps add extra memory to be sure you don't run out). Do mix the recordings during off-hours if you can.

You should also post your present server load (1 min/5 min/10 min from top right corner of top or htop) as well as what you observe as "peak" when the system passes the one minute mark and the "timed events" occur. (Peaking near 90% is sort of your limit, 100% is dangerous.)

Sorry for Hacking the thread But :

1) Enable RAMdrive or tmpfs. : How to do it and whats the advantage of it ?

2) Mix the recordings in MP3 during off hours and etc. :I have Tried it but changing parameters in cron file But qualty is pathetic. Wav format is quality if much better but sizes are too big.

Gardo has posted a couple times, as have others. The concept is to use RAM for the recording files (requires more memory, of course) to avoid massive disk i/o required for creating the recordings. then "dumping" the recordings on a regular basis, timed event, instead of constantly creating them.

so var/spool/asterisk/monitor should be a RAM drive to speed things up. then when files are moved to var/spool/asterisk/monitorDONE, it's because the conversation is over and the file is ready for processing (and it's no longer in a RAM drive).

MP3 is better than GSM, WAV is better than MP3. personal taste. look into the sox program for file conversion possibilities (it's on your server already and being used for conversions).

In terms of quality of voice, GSM is actually better than MP3, and several clients have confirmed this, you can hear more clearly with GSM, but MP3 is more universally accepted and easier to play on any computer or device so that is what is usually used.

i'm gonna have to test that thanks.