search

danswer-ai / danswer

Gen-AI Chat for Teams - Think ChatGPT if it had access to your team's unique knowledge.
https://docs.danswer.dev/
Other
10.35k stars 1.24k forks source link

The backend is currently unavailable. and all server is started. #431

Open ratel-hu opened 1 year ago

ratel-hu commented 1 year ago

image

show the notice:The backend is currently unavailable. and all server is started.

who can give me some advice

skylight-f commented 1 year ago

这个问题有没有解决方法的，我也有这个问题？

Is there a solution to this problem? I have the same problem？

GuillaumeDeWin commented 1 year ago

Same for me. Couldn't find a fix yet

ratel-hu commented 1 year ago

you should find the logs, and confirm if the api_server started

ratel-hu commented 1 year ago

in my deployment, the server is always restart. image

xhaaix commented 11 months ago

Have the same issue following the steps in the Quickstart guide. The error is different however:

  File "/usr/local/lib/python3.11/site-packages/requests/adapters.py", line 519, in send
    raise ConnectionError(e, request=request)
requests.exceptions.ConnectionError: HTTPConnectionPool(host='index', port=19071): Max retries exceeded with url: /application/v2/tenant/default/prepareandactivate (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f93b642f890>: Failed to establish a new connection: [Errno -5] No address associated with hostname'))

ERROR:    Application startup failed. Exiting.
INFO  [alembic.runtime.migration] Context impl PostgresqlImpl.
INFO  [alembic.runtime.migration] Will assume transactional DDL.
Starting Danswer Api Server

One can see in the log above that API could not connect to the index container which is a Vespa instance. As turned out the latest version (and some versions before it) of this image is broken and could not be initialized: /usr/local/bin/start-container.sh: line 67: 10 Illegal instruction /opt/vespa/bin/vespa-start-configserver

I solved the problem by downgrading to vespaengine/vespa:8.217.46 which appears to be working (for now :))

stemae commented 10 months ago

in my deployment, the server is always restart. image

works on my mac with docker, but I get the same problem when using podman. The api server is constantly restarting.. anyone any clue?

andreer commented 6 months ago

One can see in the log above that API could not connect to the index container which is a Vespa instance. As turned out the latest version (and some versions before it) of this image is broken and could not be initialized: /usr/local/bin/start-container.sh: line 67: 10 Illegal instruction /opt/vespa/bin/vespa-start-configserver

I solved the problem by downgrading to vespaengine/vespa:8.217.46 which appears to be working (for now :))

What is happening in this specific case is that Vespa requires a newer CPU - your CPU is too old to run the latest versions of Vespa, which are highly optimized for CPUs produced in the last 10 years.

See https://hub.docker.com/r/vespaengine/vespa-generic-intel-x86_64/tags for a special docker image that is slower, but works on older CPUs.

radiogena commented 4 months ago

confirmed cpu needs to be changed in proxmox to v4 and everything started as expected