cass_session_connect future takes long than 5 minutes to return

[Robin Mahony]

Hi,

So I am seeing that when I call cass_session_connect(), follow by cass_future_get_error_code(); that the future get takes longer than 5 minutes to return. I have some code that treats this as a thread deadlock, which causes crashes in my application.

This seems to happen when the local Cassandra node has just started up, and we try to connect pretty immediately after that.

Any help would be appreciated.

Cheers,

Robin M

[Michael Penick]

Are you seeing anything in the driver log (errors, warnings, etc.) when this happens?

Mike

--
You received this message because you are subscribed to the Google Groups "DataStax C++ Driver for Apache Cassandra User Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cpp-driver-us...@lists.datastax.com.

[Robin Mahony]

I haven't noticed anything, but I'll look next time it happens more thoroughly.

[Robin Mahony]

So it happened again. Here are the errors I see.

Expected, since node is coming up:

May 12 11:45:41 [ERROR] (src/connection.cpp:534:static void cass::Connection::on_connect(cass::Connector*)): Connect error 'connection refused' on host 10.96.100.241
May 12 11:45:41 [ERROR] (src/connection.cpp:813:void cass::Connection::log_error(const string&)): Host 10.96.100.241 had the following error on startup: Unable to connect

Expected, due to the way we configure our Cassandra.yaml (and unrelated in general I would think)

[WARN] (src/control_connection.cpp:96:static bool cass::ControlConnection::determine_address_for_peer_host(const cass::Address&, const cass::Value*, const cass::Value*, cass::Address*)): Found host with 'bind any' for rpc_address; using listen_address (10.96.100.243) to contact instead. If this is incorrect you should configure a specific interface for rpc_address on the server.

I think I've seen this when nodes go down while the driver is connected, so this may be expected as well

[ERROR] (src/connection.cpp:810:void cass::Connection::log_error(const string&)): Host 10.96.100.236 had the following error: Unable to decrypt data:

This I've never noticed before:

May 12 11:44:12 [WARN] (src/control_connection.cpp:211:virtual void cass::ControlConnection::on_close(cass::Connection*)): Lost control connection on host 10.96.100.236

[Michael Penick]

I've been able to reproduce the issue: https://datastax-oss.atlassian.net/browse/CPP-367

Working on a fix.

Mike

[Robin Mahony]

Awesome. Any rough idea of when this patch will be published with a release?

[Michael Penick]

There's a patch available in this pull request: https://github.com/datastax/cpp-driver/pull/288/files

This will be included with the 2.4 release in a few weeks.

Mike

On Tue, May 17, 2016 at 10:49 AM, Robin Mahony <rob...@netapp.com> wrote:

Awesome. Any rough idea of when this patch will be published with a release?

[Robin Mahony]

Ok. Assuming this patch should be safe to apply to 2.2.2?

[Michael Penick]

Yes.

On Wed, May 18, 2016 at 3:47 PM, Robin Mahony <rob...@netapp.com> wrote:

Ok. Assuming this patch should be safe to apply to 2.2.2?