Conversation
bug fix: repair fail on send large query(more than 16MB). bug fix: repair funtion _get_byte8, bit operate only can work right at less than 32.
added test for read large row added test for large insert_id
|
Thank you for your patches! It'll be even more appreciated if you can
Thanks! |
|
修改的git log的描述 |
|
Perfect! Thank you for your contribution! I'll look into your patches soon :) |
…ids) could not be properly parsed due to the lack of support for 64-bit integers in LuaJIT's BitOp library. thanks Azure Wang for the patch in github pull #6.
…t case added in the previous commit for pull #6.
|
Your test case "TEST 2: large query" in t/large.t is failing on my side. I'm getting the following error: Do you have any ideas? Is it passing on your side? Do you have any special configurations on the mysqld side? |
|
And I'm getting the same error while running your "TEST 3: large row" in t/large.t: |
|
应该是mysqld端设置的最大包大小不够大。我检查我的环境,有可能在测试脚本里加一下设置环境变量。
|
|
你确定打过补丁在跑的测试? |
|
我这边在mysqld设置max-allowed-packet=200m
|
|
我的mysqld是debian6.0上apt-get安装的mysql-server。配置文件只增加了上面说的那一项。 在 2013-3-28 上午11:26,"azure wang" azure1st@gmail.com写道:
|
|
你可以把你的mysqld信息告诉我一下,我按照你的mysqld环境在测试一下。
|
|
测试三,我这没打补丁时,得到的行数是错误的,应该是一行,得到了两行。行里的数据也是错误的。没有出现管道破裂的情况啊。很奇怪。
|
|
Hello! 2013/3/27 AzureWang notifications@github.com
What happens if you use the default mysqld setup? I think the tests should pass with the default mysqld configuration. -agentzh |
|
Hello! On Wed, Mar 27, 2013 at 7:21 PM, AzureWang notifications@github.com wrote:
Yes, of course. I was on your fork's master. Regards, |
|
Hello! 2013/3/27 AzureWang notifications@github.com
I'm using Fedora 17's mysqld (everything updated from the official yum), -agentzh |
|
我修改了测试的代码,你出现的错误是由于使用了mysqld5.5导致的,我写测试脚本的时候使用的是5.1,mysqld5.5如果发送的SQL超过max-allowed-packet(1M)就出现broken pipe, mysqld5.1会明确告诉你Got a packet bigger than 'max_allowed_packet' bytes. |
|
Okay, I get it now :) |
fix receive large packet
fix _get_byte8
fix check large query