503 Errors - The request could not be satisfied

[ytwguru]

Describe the bug
We get irregular 503 errors for dynamic and static pages.

Updated
These seem to happen when the lambda requires more than 10 seconds to run

Expected behavior
After reloading the page, we see

503 ERROR

The request could not be satisfied.

The Lambda function associated with the CloudFront distribution is invalid or doesn't have the required permissions. We can't connect to the server for this app or website at this time. There might be too much traffic or a configuration error. Try again later, or contact the app or website owner.
If you provide content to customers through CloudFront, you can find steps to troubleshoot and help prevent this error by reviewing the CloudFront documentation.

Generated by cloudfront (CloudFront)

[MBach]

Hi @ytwguru

Actually, I have the same problem. I don't know how to solve it, can you explain why you closed this issue?

[ytwguru]

@MBach. I still have the same problem, and it occurs irregularly. The issue was discussed here and I didn't want to create a duplicated issue.

I am still troubleshooting and reviewing the AWS CloudWatch Logs, but I can post an update or a PR if I get a solution.

[danielcondemarin]

@ytwguru What have you found in your CloudWatch logs? I'm keen on getting to the bottom of any 503 errors.

[ytwguru]

@danielcondemarin My intermittent errors are with the graphql server.

"errorType": "Runtime.UnhandledPromiseRejection"
"errorMessage": "Error: Network error: request to https://****.com/graphql failed, reason: Client network socket disconnected before secure TLS connection was established"

Seems like some configuration of with next and apollo.

[mertyildiran]

@danielcondemarin hi, I would like to resurrect this issue because I'm getting the exact same error:

Though I couldn't really track the Request ID, I guess it's caused by this error in my case:

{
    "errorType": "Error",
    "errorMessage": "Cannot find module './pages/article/[...slug].js'",
    "code": "MODULE_NOT_FOUND",
    "stack": [
        "Error: Cannot find module './pages/article/[...slug].js'",
        "    at Function.Module._resolveFilename (internal/modules/cjs/loader.js:636:15)",
        "    at Function.Module._load (internal/modules/cjs/loader.js:562:25)",
        "    at Module.require (internal/modules/cjs/loader.js:692:17)",
        "    at require (internal/modules/cjs/helpers.js:25:18)",
        "    at /var/task/index.js:59:18",
        "    at Generator.next (<anonymous>)",
        "    at /var/task/index.js:8:71",
        "    at new Promise (<anonymous>)",
        "    at __awaiter (/var/task/index.js:4:12)",
        "    at Runtime.exports.handler (/var/task/index.js:36:30)"
    ]
}

and a screenshot of the logs might be more meaningful:

The thighs I observed are:

• Firstly, even though I used only getStaticProps() and getStaticPaths(), the Static Site Generation (SSG) does not take place because the latency didn't make me feel like it's a static content and it's executing DB quires. While on my local machine, after the build, it's obvious that the pages are static. Probably I misconfigured something on AWS-side.
• Secondly, this error only occurs on pages and API endpoints with dynamic routing.

The project that I'm working on is closed-source so I cannot share a sample code. But I can give you some info:

• next version: 9.3.4
• serverless-next.js version: ^1.11.1-alpha.0 (development dependency)

serverless.yml:

nextApp:
  component: "serverless-next.js"

next.config.js:

module.exports = {
  target: 'serverless',
}

and I deploy it by just running serverless.

[danielcondemarin]

@mertyildiran Try version @1.11.1 (not the alpha) which includes a fix for catch all routes. That should fix the error you're seeing.
Also note that getStaticPaths and getStaticProps have still not been implemented. I have created an RFC for that : #355

[mertyildiran]

@danielcondemarin yarn add serverless-next.js -D installed 1.11.1-alpha.0 by default. Downgrading to 1.11.1 solved the issue thank you so much! 😊

[afenton90]

I'm encountering this issue with @sls-next/serverless-component@^1.17.0 still very intermittent, with no real pattern as to when the 503 appears, but I am getting the same error message as @ytwguru . Any thoughts on what could be causing this, or links to related issues @danielcondemarin ?

[danielcondemarin]

@afenton90 Can you try using the latest alpha please https://github.com/serverless-nextjs/serverless-next.js/releases/tag/%40sls-next%2Fserverless-component%401.18.0-alpha.27 ? We're about to promote that alpha to stable soon.

[campbellgoe]

Just got same 503 error with component: "@sls-next/[email protected]" although this is the first time I've set up next-auth + magic email link so it might be something wrong on my side.

[grantmontgomery]

Also experiencing the same issue when trying to request an API route with Apollo. All my other api's work in production, but it's just this one that's coming back with the same error. My version of the sls-next/serverless component is also 1.18.0, but I'll try the latest alpha release.

[a-game]

I just encountered the same problem with @sls-next/[email protected]

Error log:

{
    "errorType": "Error",
    "errorMessage": "Cannot find module '../chunks/775.js'\nRequire stack:\n- /var/task/pages/index.js\n- /var/task/default-handler-1d169422.js\n- /var/task/index.js\n- /var/runtime/UserFunction.js\n- /var/runtime/index.js",
    "code": "MODULE_NOT_FOUND",
    "requireStack": [
        "/var/task/pages/index.js",
        "/var/task/default-handler-1d169422.js",
        "/var/task/index.js",
        "/var/runtime/UserFunction.js",
        "/var/runtime/index.js"
    ],
    "stack": [
        "Error: Cannot find module '../chunks/775.js'",
        "Require stack:",
        "- /var/task/pages/index.js",
        "- /var/task/default-handler-1d169422.js",
        "- /var/task/index.js",
        "- /var/runtime/UserFunction.js",
        "- /var/runtime/index.js",
        "    at Function.Module._resolveFilename (internal/modules/cjs/loader.js:815:15)",
        "    at Function.Module._load (internal/modules/cjs/loader.js:667:27)",
        "    at Module.require (internal/modules/cjs/loader.js:887:19)",
        "    at require (internal/modules/cjs/helpers.js:74:18)",
        "    at Object.__webpack_require__.f.require (/var/task/pages/index.js:1980:28)",
        "    at /var/task/pages/index.js:1904:40",
        "    at Array.reduce (<anonymous>)",
        "    at Function.__webpack_require__.e (/var/task/pages/index.js:1903:67)",
        "    at Function.__webpack_require__.x (/var/task/pages/index.js:1996:33)",
        "    at /var/task/pages/index.js:2018:57"
    ]
}

[dphang]

@a-game That is due to newer Next.js versions changing their code-splitting, it should be fixed in the latest alpha version (note for future issues please first try to use the latest alpha as it might fix the bug you are encountering)

Also it is recommended to not comment on closed issues (unless its related and was not resolved) as it's hard for people to track except those already subscribed to that issue (And maintainers only get a single email)

[ViniciusFXavier]

Same error here

[rahuls-gfg]

I was facing this same issue, the error was happening because I have turned off the database in RDS. The backend service is unable to connect to the database.

Once the database is turned on, the app started working properly.