API hard down 500 errors now do a good job šŸ‘ fix it guys


It has been a few days with 500 errors. I see this mostly for the embedding endpoint and it is frustrating

I’m seeing the same issue.

openai.error.APIError: Internal server error {
ā€œerrorā€: {
ā€œmessageā€: ā€œInternal server errorā€,
ā€œtypeā€: ā€œauth_subrequest_errorā€,
ā€œparamā€: null,
ā€œcodeā€: ā€œinternal_errorā€
}
}

It started raising the error since a hour ago. I’m using ā€œgpt-3.5-turbo-instructā€ model.
In PST, the bay area.

Hope OpenAPI can provide some SLAs so I can use it in production env confidently.

Same here, it’s happening intermittently on my end.

You’re in the forum specifically dedicated to the API, and people are having issues working due to API server errors. This is surprising to you?

my bad, ill do better next time, sorry

Time to fire up lmdeploy on the V100 cluster…

{ā€œerrorā€:{ā€œcodeā€:502,ā€œmessageā€:ā€œBad gateway.ā€,ā€œparamā€:null,ā€œtypeā€:ā€œcf_bad_gatewayā€}}

ā€œerrorā€: {ā€œmessageā€: ā€œInternal server errorā€,ā€œtypeā€: ā€œauth_subrequest_errorā€,ā€œparamā€: null,ā€œcodeā€: ā€œinternal_errorā€}

URL:https://api.openai.com/v1/chat/completions POST Code:500 failed.
URL:https://api.openai.com/v1/chat/completions POST Code:502 failed.

are packed in a log file :dizzy_face:

:face_with_spiral_eyes:

Down for me now, consistently.

in _interpret_response_line
raise self.handle_error_response(
openai.error.APIError: Internal server error {
ā€œerrorā€: {
ā€œmessageā€: ā€œInternal server errorā€,
ā€œtypeā€: ā€œauth_subrequest_errorā€,
ā€œparamā€: null,
ā€œcodeā€: ā€œinternal_errorā€
}
}
500 {ā€˜error’: {ā€˜message’: ā€˜Internal server error’, ā€˜type’: ā€˜auth_subrequest_error’, ā€˜param’: None, ā€˜code’: ā€˜internal_error’}} {ā€˜Date’: ā€˜Fri, 20 Oct 2023 01:37:56 GMT’, ā€˜Content-Type’: ā€˜application/json; charset=utf-8’, ā€˜Content-Length’: ā€˜166’, ā€˜Connection’: ā€˜keep-alive’, ā€˜vary’: ā€˜Origin’, ā€˜x-request-id’: ā€˜dabace7d4d6ec9c96b3411eaf9ebd84’, ā€˜strict-transport-security’: ā€˜max-age=15724800; includeSubDomains’, ā€˜CF-Cache-Status’: ā€˜DYNAMIC’, ā€˜Server’: ā€˜cloudflare’, ā€˜CF-RAY’: ā€˜818d9b8be3767d7-SJC’, ā€˜alt-svc’: ā€˜h3=ā€œ:443ā€; ma=86400’}

RIP getting the same error :frowning:

ā€œerrorā€: {
ā€œmessageā€: ā€œInternal server errorā€,
ā€œtypeā€: ā€œauth_subrequest_errorā€,
ā€œparamā€: null,
ā€œcodeā€: ā€œinternal_errorā€
}

Still hard down here my servers are in NYC

It is being looked into, please be patent.

error: {
code: 502,
message: ā€˜Bad gateway.’,
param: null,
type: ā€˜cf_bad_gateway’
},

Down here in Singapore too

0_o yeaaa… failed about 500 failed calls in a row in my app… still seems to be down. I really need to build a fall back LLM into the system

Azure OpenAI has proper SLA. I’m probably going to switch if the usage grows

Would be nice to hear what is wrong and what didn’t work as a fix these guys might have some good feedback

Got the same error here

Internal server error {
ā€œerrorā€: {
ā€œmessageā€: ā€œInternal server errorā€,
ā€œtypeā€: ā€œauth_subrequest_errorā€,
ā€œparamā€: null,
ā€œcodeā€: ā€œinternal_errorā€
}
}

500 {ā€˜error’: {ā€˜message’: ā€˜Internal server error’, ā€˜type’: ā€˜auth_subrequest_error’, ā€˜param’: None, ā€˜code’: ā€˜internal_error’}}
I also got same error

Just saw first sucess

Seems now intermittent works


儽家伙