You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@nifi.apache.org by "Martin Zink (Jira)" <ji...@apache.org> on 2023/06/30 14:02:00 UTC
[jira] [Updated] (MINIFICPP-2131) GetTCP's concurrent-handler-count not threadsafe
[ https://issues.apache.org/jira/browse/MINIFICPP-2131?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Martin Zink updated MINIFICPP-2131:
-----------------------------------
Status: Patch Available (was: In Progress)
https://github.com/apache/nifi-minifi-cpp/pull/1592
> GetTCP's concurrent-handler-count not threadsafe
> ------------------------------------------------
>
> Key: MINIFICPP-2131
> URL: https://issues.apache.org/jira/browse/MINIFICPP-2131
> Project: Apache NiFi MiNiFi C++
> Issue Type: Bug
> Reporter: Martin Zink
> Assignee: Martin Zink
> Priority: Major
> Time Spent: 2h
> Remaining Estimate: 0h
>
> GetTCP utilizes his own threadpool to handle the connections, but the handlers and the futures are not properly synced, they are reliant that the worker_threads from the threadpool access the socket_ring_buffer in the same order which they are instantiated which is not a guarantee.
> With multiple endpoints (endpoint-list) and multiple concurrent handlers (concurrent-handler-count > 1) the endpoints and their respective futures could get mixed up, with unintended consequences like multiple handlers for one endpoint and none for the other or some read access violation.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)