How to configure Weaviate server write timeout in deployment yaml

Support
1

I have a Python project with Weaviate 1.8.0 to index more than 4 million vectors. During indexing I met the error:

An I/O timeout occurs when the request takes longer than the specified server-side timeout.

As mentioned in the message, I configured a client timeout:
weaviate_client = weaviate.Client(weaviate_server_url, timeout_config=(600, 600))
I also reduced the batch size.

How to configure a write timeout at server-side?
Our Weaviate runs on a Kubernetes pod with a deployment yaml file. What’s the command in deployment yaml to run Weaviate with --write-timeout=600s as a command line flag? I didn’t find the right documentation.

2

See this line in our official Helm chart.

I have a Python project with Weaviate 1.8.0

Assuming that’s not a typo, that’s a very old version. The current version (as of the time of writing this) is v1.21.3. v1.8.0 was released almos 2 years ago and a lot has improved since then. You might want to consider upgrading to the latest version :wink:

3

Hey @garance - do you mind letting us know if you saw a reference to Weaviate 1.8.0 somewhere?

We see outdated references out there from time to time so maybe we can update it for the next person :slight_smile:

4

Thank you for the answer. It’s a project created 3 years ago, we will try the latest version.

5

I have another question, if I upgrade Weaviate to the latest version, may I restore indexes of v1.8.0 in this version?