Apply the use Thread.handle_interrupt to protect query fix to the 0.3.x branch
#731
Conversation
rrosenblum
force-pushed
the
apply_592_to_3x
branch
from
a0bf33a to
a171a85
Compare
|
This is happening with one of our Rails apps in test mode. I'm running Ruby 2.2.1, Rails 4.1.14, mysql2 0.3.20, MySQL 5.6. Let me know if I could be of any help. |
|
@chrisledet, I am confident that this will fix the issue because it is the same fix that was applied to |
|
@rrosenblum Yupp, thanks for saving us the headache of searching for this bug and fixing it. |
rrosenblum
force-pushed
the
apply_592_to_3x
branch
from
a171a85 to
81632e2
Compare
|
I didn't see mention of this in the other PR or issues, but I think this will also fix the error: |
|
This bugfix is very badly needed. The fallout from this error is pretty destructive for Rails apps considering the fact that we end up with an in-use connection placed back into the pool. 👍 this PR! |
|
I haven't heard anything in a few months. @sodabrew and @brianmario what are your thoughts on this? |
|
I looked this over, and it's missing some of @tamird's changes to client.c. I'll put together an equivalent rebase in a new PR. |
|
Resolved as #751. Thanks very much for nudging me to ship this. |
This is to back port of @tamird's #592 to the
0.3.xbranch.I recently came across the issue,
This connection is still waiting for a result, try again once you have the result. I am running Ruby 2.2.4, Rails 3.2.22, Mysql2 0.3.20, and MySQL 5.6. I came across issues #99 and #566. These led me to PR #592. I saw that the fix was applied tomaster(0.4.x), but not to0.3.x. This change was able to solve the problem that I was experiencing. I was wondering why the fix was not applied to the0.3.xbranch?