Troubleshooting

Also see the Vespa FAQ and Vespa support for more help resources.

Vespa.ai and pyvespa

Both Vespa and pyvespa APIs change regularly - make sure to use the latest version of vespaengine/vespa by running docker pull vespaengine/vespa and install pyvespa.

python3 -m pip show pyvespa shows current version.

Docker Memory

pyvespa will start a Docker container with 4G memory by default - make sure Docker settings have at least this. Use the Docker Desktop settings or docker info | grep "Total Memory" to validate.

Port conflicts / Docker

Some of the notebooks run a Docker container. Make sure to stop running Docker containers before (re)running pyvespa notebooks - run docker ps and docker ps -a -q -f status=exited to list containers.

Deployment

Vespa has safeguards for incompatible deployments, and will warn with validation-override or INVALID_APPLICATION_PACKAGE in the deploy output. See validation-overrides. Most often is this due to pyvespa reusing a Docker container instance, and the fix is to list - docker ps - and remove - docker rm -f <container id> - the existing Docker containers. Alternatively, using the Docker Dashboard application. Then deploy again.

After a deployment, validate status:

Config server state: http://localhost:19071/state/v1/health
Container state: http://localhost:8080/state/v1/health

Look for "status" : { "code" : "up"} - both URLs must work before feeding or querying.

Full disk

Make sure to allocate enough disk space for Docker in Docker settings - if writes/queries fail/no results, look in the vespa.log (output in the Docker dashboard):

WARNING searchnode proton.proton.server.disk_mem_usage_filter Write operations are now blocked: 'diskLimitReached: { action: "add more content nodes", reason: "disk used (0.939172) > disk limit (0.9)", stats: { capacity: 50406772736, used: 47340617728, diskUsed: 0.939172, diskLimit: 0.9}}'

Future pyvespa versions might throw an exception in these cases. See Feed block - Vespa stops writes before the disk goes full. Add more disk / clean up, or follow the example to reconfigure for higher usage.

Check number of indexed documents

For query errors, check the number of documents indexed before debugging further: app.query(yql='select * from sources * where true).number_documents_indexed.

If this is zero, check that the deployment of the application worked, and the subsequent feeding step.

Too many open files during batch feeding

This is an OS-related issue. There are two options to solve the problem:

Reduce the number of connections via the connections parameter: with app.syncio(connections=12):.
Increase the open file limit: ulimit -n 10000. Check if the limit was increased with ulimit -Sn.

Data export

vespa visit exports data from Vespa - see Vespa CLI. Use this to validate data feeding and troubleshoot query issues.