On 2/5/2021 5:08 PM, deon wrote:
>> Really <3 Docker these days, as I can spin up/down everything I need as
>> project dependencies and spin them down when switching to something
>> else.  But I still need more than 8gb ram typically.
> 
> Wow, you must be running some large stuff on your laptop?

Nothing too crazy... a typical work project, as an example, includes
postgresql, redis, rabbitmq, an api, a couple worker services, the ui,
and may run additional integration tests...

Usually hit around 11-12gb of typical use if I'm not doing much with
the various services in the background, a couple browsers, vs code and a 
terminal window or two.


> I too develop on a MacBook Pro - and my iMac - depends where I am as 
> to which one I use.

Not really using a mac right now... work laptop is a Dell (POS) and 
Desktop is an r9-3950X.  Both currently using Windows 10, but most of
my work is under WSL2 with Ubuntu, and Docker Desktop installed via WSL2 
as well.

> I think I've only given my iMac 4GB of ram for docker to work with
> (my laptop only gives docker 2GB), and I load some large DB's in it 
> (there must be about 1-2mil records) and it performs OK, not fast I 
> admit. Maybe that's not large? :)

I have my work laptop capped to 12gb for docker, it usually hovers 
around 8-10gb or so, I'm on a loaner that only has 16gb right now, not 
sure what the plan is with my issued laptop (3rd motherboard in 2.5 
years, and now it blue screens on windows updates).


> But you are right - for developing, docker makes things pretty easy. 
> I use "syncthing" to keep my laptop and imac development directory 
> in sync, so when I switch between one or the other, I just kill the 
> containers on my laptop and restart them on the imac and keep going 
> (or visa-versa).

Mostly not doing much syncing... just relying on source control (git) 
for the most part.

> When I'm finished with a feature, I push the code to my github, 
> which (depending on the project), builds the container that ends 
> up on the live server. With a Dockerfile that is consistent 
> between environments there is rarely any localisation issues with 
> deployment...

I've got a few things like that... work projects have a PR process in 
place, then goes through a Dev->UAT->Prod release cycle.  Other
projects have different release cycles as different states/counties have 
different limitations on when releases can happen.  Been a long, uphill 
battle, but most of our software is now released via CI/CD deployment 
agents/pipelines, and a lot of the testing is now containerized.

The application I'm spending most of my time is docker-compose for 
local/testing and kubernetes release env.
-- 
Michael J. Ryan (tracker1)
+o roughneckbbs.com
---
 þ Synchronet þ Roughneck BBS - roughneckbbs.com
--- SBBSecho 3.12-Linux