My prod.log file is getting very large, so I want to use the rotating_file option to create one file per day (max 14). It works, everyday a new file is generated but the file prod.log still exists and logs all new messages. Is it possible to prevent the generation of prod.log file? I use Symfony 2.8, monolog 1.23.0 and the default configuration in config.yml:
monolog:
use_microseconds: false
handlers:
main:
type: rotating_file
path: '%kernel.logs_dir%/%kernel.environment%.log'
level: error
max_files: 14
Problem solved. I disabled full monolog configuration, cleared both caches, reenabled the configuration, cleared cache again, now it works. Strange cache problem, seems both configurations were used.
Related
I want to log only the deprecated warnings to a new file. But I can't see how to achieve that with monolog. Is there a custom configuration?
Thanks in advance!
All Deprecated Message are logged As INFO level, php Channels so if You try this configuration All deprecation message will be logged in one file
monolog:
handlers:
security:
level: INFO
type: stream
path: '%kernel.logs_dir%/deprecated.log'
channels: [php]
Background: we're using Symfony 3.1 + Doctrine 2.5.5 + symfony doctrine bundle.
While trying to enable second level caching for our entities, we have encountered the following issue. If we use NONSTRICT_READ_WRITE, everything works fine. However, when we tried to use READ_WRITE, we got the following error
0)
Type error: Argument 2 passed to Doctrine\ORM\Cache\Persister\Entity\ReadWriteCachedEntityPersister::__construct() must be an instance of Doctrine\ORM\Cache\ConcurrentRegion, instance of Doctrine\ORM\Cache\Region\DefaultRegion given, called in vendor/doctrine/orm/lib/Doctrine/ORM/Cache/DefaultCacheFactory.php on line 133
As far as I understood, we need to actually implement our own version of ConcurrentRegion and CacheFactory to make it work (FileLockRegion does not suit us due to its usage of file system to handle cache locks). So I wrote those implementations, but the main issue now lies in following: I cannot find where to put my custom factory class' name in the configuration. We have tried the following locations in config:
1)
doctrine:
orm:
auto_generate_proxy_classes: "%kernel.debug%"
default_entity_manager: master
second_level_cache:
default_cache_factory:
class: AppBundle\Cache\MyCacheFactory
This fails due to
Unrecognized option "second_level_cache" under "doctrine.orm"
even though in our other project using Symfony 2.8 option "second_level_cache" does not throw any errors.
So we went to doctrine bundle code and found the following node description (vendor/doctrine/doctrine-bundle/DependencyInjection/Configuration.php:492)
->arrayNode('second_level_cache')
->children()
->append($this->getOrmCacheDriverNode('region_cache_driver'))
->scalarNode('region_lock_lifetime')->defaultValue(60)->end()
->booleanNode('log_enabled')->defaultValue($this->debug)->end()
->scalarNode('region_lifetime')->defaultValue(0)->end()
->booleanNode('enabled')->defaultValue(true)->end()
->scalarNode('factory')->end()
->end()
So we decided we should try this config in our master entity manager section:
2)
second_level_cache:
region_cache_driver:
type: memcache
enabled: true
log_enabled: true
factory: AppBundle\Cache\MyCacheFactory
regions:
hour_region:
lifetime: 3600
However, even though this config is considered valid, when we actually try to access the entity with configured caching, we get the error 0), which makes us think that this option is being ignored by doctrine/symfony.
Is there any way to do it via .yml config at all? Doctrine docs only propose to implement CacheFactory and provide a PHP code example, but it's still quite unclear where should this PHP code go, even if we decide to abandon the idea of putting our class in .yml config and go the PHP way.
Use type - filelock for configurate FilelockRegion
regions:
default:
cache_driver:
type: service
id: 'Doctrine\Common\Cache\RedisCache'
lifetime: 3600
type: filelock
I'm getting the following exception:
InvalidConfigurationException: The child node "clients" at path "snc_redis" must be configured.
I use Symfony 2.3 SncRedisBundle 1.1.x-dev
In config.yml I have
clients:
default:
type: predis
alias: default
dsn: redis://localhost
logging: %kernel.debug%
And I've followed the docs here for installation...
Any idea what would cause this ?
The bundle is looking for configuration values under snc_redis root (as you can see here), which you did not provide.
Now, all this wouldn't normally be a problem, but the clients child node is mandatory (as you can see here), which causes the error above.
Simply make the root configuration node in your config.yml to be snc_redis and the problem will go away. You probably simply missed it when doing copy-paste.
I'm trying to optimize my project logs (Because my log got 3Gb at now) so that when something goes wrong the server send me an email with the details of the error.
I would like to appear in the log only major errors, such as 500 errors, errors that affect the proper functioning of the project.
I have looked at the documentation of the bundle of monolog on the official Symfony2 but I have not been at all clear.
(http://symfony.com/doc/current/reference/configuration/monolog.html)
Could someone tell me how to get this?
We defined
monolog:
handlers:
main:
action_level: error
in our production environment.
Of course you have to check which handler (in our case: 'main') you need to adapt but by changing the action_level to 'error' you get rid of all the debug/info statements in your log and only level 'error' is shown.
Please check your Swift part:
You have level: 'debug' which is obviously not corresponding with your requirement. You should use 'error' or 'critical' instead. See http://symfony.com/doc/current/cookbook/logging/monolog_email.html as well.
Generally it would be helpful to know what kind of log is producing too much and what kind of information? (e.g. we put Doctrine to a different monolog channel to get rid of it in our main log).
Here is my config for monolog.
monolog:
handlers:
main:
action_level: error
console:
type: console
bubble: false
mail:
type: fingers_crossed
action_level: critical
handler: buffered
buffered:
type: buffer
handler: swift
swift:
type: swift_mailer
from_email: his#email.com
to_email: my#email.com
subject: Critical error spotted
level: debug
I cannot get the symfony2 configuration to correctly overwrite values from other config-files. Here is the problem:
I have a new environment "staging" where I want to use most of the stuff from config_prod.yml but have another logging level (I want it to be as it is in development, simply logging everything to a file). Here are the config stuff I use:
config_prod.yml:
imports:
- { resource: config.yml }
monolog:
handlers:
main:
type: fingers_crossed
action_level: error
handler: nested
nested:
type: stream
path: %kernel.logs_dir%/%kernel.environment%.log
level: debug
config_staging.yml:
imports:
- { resource: config_prod.yml }
monolog:
handlers:
main:
type: stream
path: %kernel.logs_dir%/%kernel.environment%.log
level: debug
nested: ~
From my point of view, the nested logger is now null and the main logs to the given file. What really happens is that he logs every message twice! The same happens when I use this for the config_staging.yml:
imports:
- { resource: config_prod.yml }
monolog:
handlers:
main:
type: stream
path: %kernel.logs_dir%/%kernel.environment%.log
level: debug
handler: ~
nested: ~
I found a workaround, setting the action_level of the main handler to debug and leaving everything else as is, but I don't like this solution. There must be a way to overwrite config stuff so I only have the main monolog handler.
Pretty much one year later I now have an understanding of what's happening and how to prevent it:
The nested handler is befilled with the configuration from the config.yml and when parsing the config_staging.yml, the yaml component does not overwrite the whole hashmap and set the value to null but tries to merge both, resulting in the same array as before.
There is a type called null which can be used ot overwrite any Logger. It does nothing and is therefor suitable for this use case:
monolog:
handlers:
main:
type: stream
path: %kernel.logs_dir%/%kernel.environment%.log
level: debug
handler: ~
nested: ~
type: null
Another solution would be to not configure any logging in the config.yml but only in the specific environment configs like config_prod.yml and so on.
Check that you don't have any repeated keys in the _staging config file -- the second one would override the first, with the net result that the first is ignored.
If you want to alter a collection by removing an element you will have to create an intermediate YAML file (importing the base) setting the collection to "null" and re-adding all required collection elements in a file which in turn imports the intermediate YAML file.
You can not simply overwrite a collection. New elements will get added but you can't remove existing ones except by the workaround described.