can't start new thread
Hello,
I have two couchbase servers in a cluster running 1.8.1. My client is a python 0.6 which periodically tries to connect to each server in the cluster for health check purposes and perform a key set/get. The bucket is a couchbase bucket.
The health script runs correctly for a while (about 30 min checking connections every 10sec). Then, I get "can't start new thread" when I try to connect.
The high level client code:
from couchbase.couchbaseclient import VBucketAwareCouchbaseClient as couchbaseClient
mydb = coucbaseClient(...)
# perform set and get
mydb.done()
I appreciate any help with this.
Thanks,
Thanks for response. Is the unified client at https://github.com/couchbase/couchbase-python-client/downloads?
No, sorry. By "unified client" I mean the 'Couchbase' class (formerly known as 'Server'). You can import it via...
from couchbase import Couchbase
It's "unified" in that it's handling both the HTTP and memcached APIs, so the code that uses it is much simpler and more focused as you'll see on the Python "next" page:
http://www.couchbase.com/develop/python/next
Hi,
I tried the unified client (Server) and it behaves the same way. A thread is created each time Server or VbucketAwareClient is invoked and the thread is never removed/stopped (even when done method in VbucketAwareClient is invoked). Eventually after about 43 calls (in my environment) it causes the error.
Thanks,
Would you mind giving the unified client a try in this case as well (per my suggestion in your other post).
Feel free to file bugs as well at our Jira for the Python SDK. It's much easier to manage issues, logs, etc via Jira than the forums.
Thanks, fm.