I tested running -j without a process limit and while it was successful
building 13 on my primary desktop (8g ram 4 core machine with ssd), it
resulted in a load average over 40, sluggish response, and ran slightly
slower than -j$(nproc). Here is my results:
1m32.204s => make -j
1m27.274s =>
On Wed, Jun 08, 2016 at 08:58:00PM +0200, Alexander Traud wrote:
> On Ubuntu 16.04 LTS, when I build Asterisk 13.9.1 via
>
>
> wget downloads.asterisk.org/pub/telephony/asterisk/asterisk-13-current.tar.gz
> tar zxf asterisk-1*.tar.gz
> sudo apt install libssl-dev libncurses-dev libnewt-dev
This is an issue in make not in asterisk so there is no point in reporting
it in the asterisk bug tracker. I tried it on mine and it started spewing
fork errors. I have 8 cores so normally do "make -j 8" with great results.
>From the man page for make: "If the -j option is given without an
I would suspect that to be a problem with make dealing successfully with
the omitted number. I use this all the time without incident:
make -j$(nproc)
On Wed, Jun 8, 2016 at 1:58 PM, Alexander Traud
wrote:
> On Ubuntu 16.04 LTS, when I build Asterisk 13.9.1 via
>
>
On Wed, Jun 8, 2016 at 1:58 PM, Alexander Traud
wrote:
> On Ubuntu 16.04 LTS, when I build Asterisk 13.9.1 via
>
>
> wget downloads.asterisk.org/pub/telephony/asterisk/asterisk-13-current.tar.gz
> tar zxf asterisk-1*.tar.gz
> sudo apt install libssl-dev libncurses-dev
On Ubuntu 16.04 LTS, when I build Asterisk 13.9.1 via
wget downloads.asterisk.org/pub/telephony/asterisk/asterisk-13-current.tar.gz
tar zxf asterisk-1*.tar.gz
sudo apt install libssl-dev libncurses-dev libnewt-dev libxml2-dev
libsqlite3-dev uuid-dev libjansson-dev libedit-dev