[
https://issues.apache.org/jira/browse/TS-4612?focusedWorklogId=26015&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-26015
]
ASF GitHub Bot logged work on TS-4612:
--------------------------------------
Author: ASF GitHub Bot
Created on: 26/Jul/16 16:26
Start Date: 26/Jul/16 16:26
Worklog Time Spent: 10m
Work Description: Github user oknet commented on the issue:
https://github.com/apache/trafficserver/pull/771
@bryancall to review the codes of ClusterHandlerBase.cc
```
979 thread = eventProcessor.eventthread[ET_CLUSTER][id %
num_of_cluster_threads];
980 if (net_vc->thread == thread) {
981 cluster_connect_state = CLCON_CONN_BIND_OK;
982 break;
983 } else {
984 cluster_connect_state =
ClusterHandler::CLCON_CONN_BIND_CLEAR;
985 }
```
From my understanding of the code, the codes from L1007 to L1033 means:
1. the net_vc bind to unmatched thread
2. It should unbind(L1012-1022) and then rebind(L1024-L1027) to another
thread
I think the unbind include remove vc from cop_list here.
because the InactivityCop only push the vc->thread == this_ethread() into
cop_list, but the vc->thread set to NULL in L1014.
```
// Copy the list and use pop() to catch any closes caused by callbacks.
forl_LL(UnixNetVConnection, vc, nh.open_list)
{
if (vc->thread == this_ethread()) {
nh.cop_list.push(vc);
}
}
```
Issue Time Tracking
-------------------
Worklog Id: (was: 26015)
Time Spent: 3h 20m (was: 3h 10m)
> Proposal: InactivityCop Optimize
> --------------------------------
>
> Key: TS-4612
> URL: https://issues.apache.org/jira/browse/TS-4612
> Project: Traffic Server
> Issue Type: Bug
> Components: Core, Network
> Reporter: Oknet Xu
> Fix For: sometime
>
> Time Spent: 3h 20m
> Remaining Estimate: 0h
>
> By review the processing of InactivityCop::check_inactivity():
> 1. get all local vc from open_list
> 2. put them into cop_list
> 3. check every vc in cop_list if it is already timeouted
> 4. callback vc->handleEvent to close vc if it is timeout
> InactivityCop and NetHandler share one mutex.
> InactivityCop runs every second, NetHandler runs every 10ms, that means
> Nethandler runs 100 times until next InactivityCop runs.
> if one vc has read/write in a Nethandler call, it is won't be timeout in the
> next InactivityCop run.
> Thus, if the vc has read/write in Nethandler, we move it out of cop-list then
> the InactivityCop runs would get better performace.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
