[
https://issues.apache.org/jira/browse/TS-4747?focusedWorklogId=26857&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-26857
]
ASF GitHub Bot logged work on TS-4747:
--------------------------------------
Author: ASF GitHub Bot
Created on: 23/Aug/16 01:22
Start Date: 23/Aug/16 01:22
Worklog Time Spent: 10m
Work Description: GitHub user keith2008 opened a pull request:
https://github.com/apache/trafficserver/pull/903
TS-4747: if the connection of parent is notalive, not make the parent host
down,which will select the the unavailablehost again
if the connection of parent is notalive, not make the parent host
down,which will select the the unavailablehost again
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/keith2008/trafficserver TS-4747
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/trafficserver/pull/903.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #903
----
commit 8a76d21f5194cb31e514860319b8fcae2d2648be
Author: keith2008 <[email protected]>
Date: 2016-08-23T00:42:30Z
TS-4743: parent use consistent_hash Strategy may cause crash while first
parent is not set
commit 3deb52397b047455ba8aba3e5132e3daafa980ad
Author: keith2008 <[email protected]>
Date: 2016-08-23T00:51:44Z
TS-4744: ParentConsistentHash::selectParent may select the unavailable
parent
commit 49beaa5b5fd712c1d8383b081fc8479ec680ae0a
Author: keith2008 <[email protected]>
Date: 2016-08-23T00:58:54Z
TS-4745: pRecord.failCount not init in ParentRecord::ProcessParents
commit 1159f47044e7add14ccaa84bf110895e244599b0
Author: keith2008 <[email protected]>
Date: 2016-08-23T01:12:22Z
TS-4746: ParentRecord *secondary_parents malloc, but no place free,which
will cause memery leak
commit 46671f91f337cd28dbc084dedde903382e3b3c38
Author: keith2008 <[email protected]>
Date: 2016-08-23T01:17:40Z
TS-4747: if the connection of parent is notalive, not make the parent host
down,which will select the the unavailablehost again
----
Issue Time Tracking
-------------------
Worklog Id: (was: 26857)
Time Spent: 10m
Remaining Estimate: 0h
> if the connection of parent is not alive, not make the parent host down,which
> will select the the unavailable host again
> ------------------------------------------------------------------------------------------------------------------------
>
> Key: TS-4747
> URL: https://issues.apache.org/jira/browse/TS-4747
> Project: Traffic Server
> Issue Type: Improvement
> Reporter: xiangdong chen
> Assignee: John Rushford
> Fix For: 7.0.0
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> the parent.config is like this:
> dest_domain=www.cxdtest.com parent=“dnstest1.com:80”;round_bin=strict
> and the dnstest1.com contains two ip:
> 192.168.1.1 （but it was down）
> 192.168.1.2
> if the request go to the parent, and the parent is domain,which contains
> multi ip. if one ip is down,but we not make the host down ,if the next
> request is comming, it will choice this ip again. we should make the host
> down.
> fix code on HttpTransact::handle_response_from_parent(State *s)
> default: {
> LookingUp_t next_lookup = UNDEFINED_LOOKUP;
> DebugTxn("http_trans", "[hrfp] connection not alive");
> s->state_machine->do_hostdb_update_if_necessary();//added by xdchen, make
> the host down,line:3606
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
