Tarantool development patches archive
 help / color / mirror / Atom feed
From: Vladimir Davydov <vdavydov.dev@gmail.com>
To: Serge Petrenko <sergepetrenko@tarantool.org>
Cc: tarantool-patches@freelists.org
Subject: Re: [PATCH v2] box: allow vinyl_memory set to 0 in config
Date: Wed, 4 Jul 2018 16:37:35 +0300	[thread overview]
Message-ID: <20180704133735.2eplki6u4ndkekzr@esperanza> (raw)
In-Reply-To: <20180704132348.74595-1-sergepetrenko@tarantool.org>

On Wed, Jul 04, 2018 at 04:23:48PM +0300, Serge Petrenko wrote:
> Sorry, forgot to post diff.

All extra information that you want to get to the reviewer should go
after --- (where you put branch and change log). This way it will be
ignored by `git am` (we don't use it now, but people in other projects
do and we might want to start using it instead of `git merge` one day).

> 
> In 1.9 it was possible to have a vinylless configuration with
> vinyl_memory=0, allow to do this in 1.10 by adjusting sanity
> checks for vinyl_memory and memtx_memory. Now banning only
> negative values.
> memtx_memory check was changed for consistency, trying to
> set memtx_memory to 0 fails anyways.
> Also added a test to check that vinyl_memory can actually
> be set to 0.
> 
> Closes: #3468
> ---
> https://github.com/tarantool/tarantool/tree/sergepetrenko/gh-3468-allow-0-vinyl-memory
> https://github.com/tarantool/tarantool/issues/3468
> 
> Changes in v2:
> - added a test case to check vinyl_memory=0 is legal.

> diff --git a/test/box-tap/cfg.test.lua b/test/box-tap/cfg.test.lua
> index b93a21f35..764fade93 100755
> --- a/test/box-tap/cfg.test.lua
> +++ b/test/box-tap/cfg.test.lua
> @@ -6,7 +6,7 @@ local socket = require('socket')
>  local fio = require('fio')
>  local uuid = require('uuid')
>  local msgpack = require('msgpack')
> -test:plan(91)
> +test:plan(93)
>  
>  --------------------------------------------------------------------------------
>  -- Invalid values
> @@ -82,9 +82,15 @@ os.execute("rm -rf vinyl")
>  box.cfg{
>      log="tarantool.log",
>      memtx_memory=104857600,
> +    vinyl_memory=0,
>      wal_mode = "", -- "" means default value
>  }
>  
> +status, result = pcall(testfun)
> +test:ok(status  and result == "table", "allow vinyl_memory set to 0 in box.cfg")
> +test:is(box.cfg.vinyl_memory, 0, "vinyl_memory actually set to 0")
> +
> +box.cfg{vinyl_memory=128 * 1024 * 1024}
>  -- gh-678: vinyl engine creates vinyl dir with empty 'snapshot' file
>  test:isnil(io.open("vinyl", 'r'), 'vinyl_dir is not auto-created')

That's not exactly what I meant. There's an easier way to test initial
box configuration:

  code = [[ box.cfg{ vinyl_memory=0 } ]]
  test:is(run_script(code), 0, 'vinyl_memory is 0')

(there are plenty of similar test cases in box-tap/cfg.test.lua)

Also, when you add a test case, please write a brief description with a
reference to the github ticket you're fixing, e.g.

(also from box-tap/cfg.test.lua)
> -- gh-2664: vinyl_dir is checked on the first use
> code = [[ box.cfg{ vinyl_dir='invalid' } ]]
> test:is(run_script(code), 0, 'vinyl_dir is invalid')

  reply	other threads:[~2018-07-04 13:37 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-04 13:23 Serge Petrenko
2018-07-04 13:37 ` Vladimir Davydov [this message]
2018-07-04 14:11   ` Sergey Petrenko
2018-07-04 14:23     ` Vladimir Davydov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180704133735.2eplki6u4ndkekzr@esperanza \
    --to=vdavydov.dev@gmail.com \
    --cc=sergepetrenko@tarantool.org \
    --cc=tarantool-patches@freelists.org \
    --subject='Re: [PATCH v2] box: allow vinyl_memory set to 0 in config' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox