Hi, We are seeing a really bizarre problem where we frequently get errors saying a class cannot be found. It's often as part of a Symfony Command, but not always. It isn't unique to one specific class, it covers lots of classes and never the same ones. e.g. here's a couple of different snippets of stack traces: Error: Class "Doctrine\Common\Collections\ArrayCollection" not found#43 /vendor/doctrine/orm/src/AbstractQuery.php(176): Doctrine\ORM\AbstractQuery::__construct#42 /vendor/doctrine/orm/src/EntityManager.php(337): Doctrine\ORM\EntityManager::createQuery#41 /var/cache/prod/ContainerFgpM7ZD/EntityManager_9a5be93.php(65): ContainerFgpM7ZD\EntityManager_9a5be93::createQuery#40 /src/Nora/ReferralBundle/Entity/Repository/MerchantRepository.php(427): Nora\ReferralBundle\Entity\Repository\MerchantRepository::findByPartnerCode...
Error: Class "Symfony\Component\Validator\Validator\RecursiveValidator" not found#35 /vendor/symfony/symfony/src/Symfony/Component/Validator/ValidatorBuilder.php(423): Symfony\Component\Validator\ValidatorBuilder::getValidator#34 /var/cache/prod/ContainerFgpM7ZD/AppKernelProdContainer.php(991): ContainerFgpM7ZD\AppKernelProdContainer::get_Container_Private_ValidatorService#33 /var/cache/prod/ContainerFgpM7ZD/getIntegrationLogsPersister2Service.php(30): ContainerFgpM7ZD\getIntegrationLogsPersister2Service::do#32 /var/cache/prod/ContainerFgpM7ZD/AppKernelProdContainer.php(582): ContainerFgpM7ZD\AppKernelProdContainer::load#31 /var/cache/prod/ContainerFgpM7ZD/getBufferedIntegrationLogsPersisterService.php(25): ContainerFgpM7ZD\getBufferedIntegrationLogsPersisterService::do#30 /var/cache/prod/ContainerFgpM7ZD/AppKernelProdContainer.php(582): ContainerFgpM7ZD\AppKernelProdContainer::load#29 /var/cache/prod/ContainerFgpM7ZD/AppKernelProdContainer.php(2027): ContainerFgpM7ZD\AppKernelProdContainer::{closure:ContainerFgpM7ZD\AppKernelProdContainer::getEventDispatcherService():2026}....
We're running PHP 8.4.6, but have seen it on PHP 8.2.28 too. We're using Symfony 5.4.47. We can't reproduce the problem locally. The really confusing thing is that this command, on this machine, may have run multiple times before the failure occurs. The machines are running on AWS Elastic Container Service and so are self scaling up and down regularly (i.e. we're regularly replacing the physical VM). I had assumed it might be something to do with when the instance first starts, but it doesn't appear to be the case - it can be after multiple successful runs it fails. I am struggling to know where to begin - does anyone have any suggestions on how I'd figure this one out? Thanks,
Ed |