[
https://issues.apache.org/jira/browse/TS-576?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12971210#action_12971210
]
Zhao Yongming commented on TS-576:
----------------------------------
from my testing, the cluster will be broken after this patch:
[Dec 14 17:32:53.291] Manager {139827441649440} NOTE:
[LocalManager::pollMgmtProcessServer] New process connecting fd '16'
[Dec 14 17:32:53.291] Manager {139827441649440} NOTE: [Alarms::signalAlarm]
Server Process born
[Dec 14 17:32:53.325] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:32:53.325] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:32:54.319] {47886601538368} STATUS: opened
/var/log/trafficserver/diags.log
[Dec 14 17:32:54.320] {47886601538368} NOTE: updated diags config
[Dec 14 17:32:54.321] Server {47886601538368} NOTE: cache clustering enabled
[Dec 14 17:32:54.349] Server {47886601538368} DEBUG: (cluster_note)
[Machine::Machine] Cluster IP addr: 10.62.163.181
[Dec 14 17:32:54.350] Server {47886601538368} NOTE: cache clustering enabled
[Dec 14 17:32:54.420] Server {47886631024400} NOTE: cache enabled
[Dec 14 17:32:54.589] Server {47886601538368} NOTE: logging initialized[7],
logging_mode = 3
[Dec 14 17:32:54.598] Server {47886601538368} NOTE: traffic server running
[Dec 14 17:32:55.287] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:32:55.287] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:32:56.300] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:32:56.300] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:32:57.313] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:32:57.313] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:32:58.326] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:32:58.326] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:33:00.288] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:33:00.288] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:33:01.310] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:33:01.310] Manager {139827375253264} ERROR: (last system error 0:
Success)
[Dec 14 17:33:02.330] Manager {139827375253264} ERROR:
[ClusterCom::handleMultiCastMessage] Invalid message-line(838) '0:1: 0'
[Dec 14 17:33:02.330] Manager {139827375253264} ERROR: (last system error 0:
Success)
and cluster can not setup with these errors flooding.
> Remove 'stats' from RecordsConfig.cc ?
> --------------------------------------
>
> Key: TS-576
> URL: https://issues.apache.org/jira/browse/TS-576
> Project: Traffic Server
> Issue Type: Bug
> Components: Stats
> Reporter: Leif Hedstrom
> Fix For: 2.1.5
>
> Attachments: TS-576.diff
>
>
> I think the "registration" of stats in RecordsConfig.cc is unecessary, and
> confusing. Should we just nuke it? Obviously make sure all those stats are
> registered with the appropriate modules (which I believe they generally are).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
