asterisk users - Mar 2011 - Why shouldn't I use 1.8?

Now that we've hashed out some thoughts on the most stable version of
asterisk, I'd like to ask the question as to why I should NOT use 1.8? What
are specific reasons? For instance a few days back I was speaking with James at
Rhino Equipment. He said that he has "no real data" on why I
shouldn't use 1.8. They just follow a practice of not jumping on the newest
version.

But I would like specific reasons why I shouldn't use 1.8 in a production
environment if anyone has some?

Thanks for your feedback!
-
Doug Mortensen
Network Consultant
Impala Networks Inc
CCNA, MCSA, Security+, A+
Linux+, Network+, Server+
.
www.impalanetworks.com
P: (505) 327-7300
F: (505) 327-7545
.

On Fri, Mar 25, 2011 at 7:36 AM, Douglas Mortensen
<doug at impalanetworks.com> wrote:
>
> But I would like specific reasons why I shouldn't use 1.8 in a
production environment if anyone has some?
That is a loaded question, in that no two environments are likely to
be the same.  Some bugs are major issues for < 1% of the install base
and take time to get merged into the code base.  You should read
through the open issues for the 1.8 branch and see if there are any
show stoppers for your environment.  If not, try it in the lab and
validate that it works for you.

Check out https://issues.asterisk.org

For my environment specifically, this issue is currently preventing me
from migrating from 1.6.2:
 - 18818 [patch] Crashing when using local channels and realtime on asterisk

There are a lot of benefits to the 1.8 branch (Long term support,
Called party id, Multicast RTP, etc) but only you can say if it will
work with your configuration in your environment.

-Jonathan

----------------------------------------
 From: "Jonathan Thurman" <jonathan at thurmantech.com>
Sent: Friday, March 25, 2011 11:06 AM
To: "Asterisk Users Mailing List - Non-Commercial Discussion" 
<asterisk-users at lists.digium.com>
Subject: Re: [asterisk-users] Why shouldn't I use 1.8?

On Fri, Mar 25, 2011 at 7:36 AM, Douglas Mortensen
<doug at impalanetworks.com> wrote:
>
> But I would like specific reasons why I shouldn't use 1.8 in a
production
environment if anyone has some?

That is a loaded question, in that no two environments are likely to
be the same. Some bugs are major issues for < 1% of the install base
and take time to get merged into the code base. You should read
through the open issues for the 1.8 branch and see if there are any
show stoppers for your environment. If not, try it in the lab and
validate that it works for you.

Check out https://issues.asterisk.org

For my environment specifically, this issue is currently preventing me
from migrating from 1.6.2:
- 18818 [patch] Crashing when using local channels and realtime on 
asterisk

There are a lot of benefits to the 1.8 branch (Long term support,
Called party id, Multicast RTP, etc) but only you can say if it will
work with your configuration in your environment.

-Jonathan

--

Doug

I agree with Jonathan. I have moved all but one of our production switches 
to 1.8 the only thing holding me back is a minor bug so I have to keep the 
1.6.2 box around until that patch is released into the 1.8 branches. When 
that is done I will no longer be on the 1.6.  I have over 98% of our load 
on the 1.8 switches and we are doing multi tenant pbx hosting and sip 
trunking.

A point of note I just turned down my last 1.4 box 2 weeks ago. It was not 
because it was not working but because I need more volume and 1.8 on the 
new hardware meets that need and I get the bonus of not having to support 
three versions of asterisk now. It is very likely that most of the time I 
will have at least two versions in production at a time. This is so I can 
offer the newest features with a stable build and I can offer a more long 
term support for the customers that the newest features are not as 
important. Most of my switch hardware has a planned 4 year life span. The 
better asterisk gets the longer I can stretch that investment. My 
recommendation is if 1.8 does not have any bugs that are issues for you try 
1.8 out of the gate and test, test, test offer feed back from your testing 
and the bugs will get fixed. 

I would not spend to much time worrying spend more time doing. 

Good luck
Bryant
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.digium.com/pipermail/asterisk-users/attachments/20110325/2344845c/attachment.htm>

On 25/03/11 14:36, Douglas Mortensen wrote:
> Now that we've hashed out some thoughts on the most stable version of
asterisk, I'd like to ask the question as to why I should NOT use 1.8? What
are specific reasons? For instance a few days back I was speaking with James at
Rhino Equipment. He said that he has "no real data" on why I
shouldn't use 1.8. They just follow a practice of not jumping on the newest
version.
>
I agree with what Jonathan also said in this thread but that is also a 
good enough reason on it's own.  Data doesn't yet exist to say whether 
it's stable enough.  I like to err on the side of caution with phone 
systems as they cost lots of money when they go wrong!