introduce unrecoverable error type, and logs to show Fluentd process/worker started #1359
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…worker started Currently, unrecoverable errors (which serverengine doesn't try to restart workers about) are only configuration errors. But some other situation cannot be retried. For example: * per-worker root directory (will be introduced next) exists but it is a file * failed to bind an address/port in server plugin helper * failed to read a file from in_tail by permission error * or many others We are using ConfigError for such cases, but it's nonsense in fact. This change also adds unrecoverable cases for scripting errors while loading plugin files. Fluentd is now restarting for such cases, but it's just wrong behavior.
|
Actually, the biggest improvement in this change is testing about:
|
|
@repeatedly could you review this change? |
repeatedly
reviewed
Dec 9, 2016
| def main_process(&block) | ||
| Process.setproctitle("worker:#{@process_name}") if @process_name | ||
|
|
||
| configuration_error = false | ||
| unrecoverable_error = false | ||
|
|
||
| begin | ||
| block.call | ||
| rescue Fluent::ConfigError |
There was a problem hiding this comment.
Use Fluent::ConfigError => e and replace $! with e.
repeatedly
reviewed
Dec 9, 2016
| log.debug_backtrace | ||
| end | ||
| unrecoverable_error = true | ||
| rescue Fluent::UnrecoverableError |
repeatedly
reviewed
Dec 9, 2016
| log.error_backtrace | ||
| end | ||
| unrecoverable_error = true | ||
| rescue ScriptError # LoadError, NotImplementedError, SyntaxError |
repeatedly
reviewed
Dec 9, 2016
|
|
||
| sub_test_case 'configuration to load plugin which raises unrecoverable error in #start' do | ||
| test 'failed to start' do | ||
| script = "require 'fluent/plugin/input'\n" |
There was a problem hiding this comment.
Writing ruby code in here document breaks indentation on my environment... :P
|
Others are LGTM and adding start test sis awesome. |
tagomoris
force-pushed
the
introduce-unrecoverable-errors
branch
18 times, most recently
from
fa4f46a to
1abe61a
Compare
tagomoris
force-pushed
the
introduce-unrecoverable-errors
branch
from
1abe61a to
af43194
Compare
Ruby 2.1/2.2 executes worker process via shell (command prompt) with previous code. Ruby 2.3 executes in different way (without shell), but we need to support 2.1/2.2. This fix forces to execute worker processes without shell.
tagomoris
force-pushed
the
introduce-unrecoverable-errors
branch
5 times, most recently
from
c08a2c8 to
cb3b3a0
Compare
Especially on Windows, test code cannot find Fluentd process id from the result of spawn, because "exec" system call cannot be used on Windows, and we got the process id of "bundle exec" by "io.pid". It is also useful to investigate supervisor-worker relation from supervisor pid and ppid of workers.
tagomoris
force-pushed
the
introduce-unrecoverable-errors
branch
from
5eb053e to
17a4880
Compare
|
I added some fixes about executing worker processes and showing process ids in log. |
|
LGTM and good fix for windows environment 👍 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Currently, unrecoverable errors (which serverengine doesn't try to restart workers about) are only configuration errors.
But some other situation cannot be retried. For example:
We are using ConfigError for such cases, but it's nonsense in fact.
This change also adds unrecoverable cases for scripting errors while loading plugin files.
Fluentd is now restarting for such cases, but it's just wrong behavior.