[jira] [Commented] (AMQ-6936) Dead loop on log file reading

2018-03-23 Thread Rural Hunter (JIRA) 

[ 
https://issues.apache.org/jira/browse/AMQ-6936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16411151#comment-16411151
 ] 

Rural Hunter commented on AMQ-6936:
---

OK, that's nice, thanks!

> Dead loop on log file reading
> -
>
> Key: AMQ-6936
> URL: https://issues.apache.org/jira/browse/AMQ-6936
> Project: ActiveMQ
>  Issue Type: Bug
>  Components: KahaDB
>Affects Versions: 5.13.4
>Reporter: Rural Hunter
>Priority: Critical
>
> We restarted our activemq instance(5.13.4) but found it hung on recovering 
> data. I got the data and ran a debug and found these:
>  # The dead loop is in Journal.getNextLocation(Location location)
>  # I added log in the method and this is I got:
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554239
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554330
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554330: offset=33554330, 
> size=28, type=2
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554358: offset=33554358, 
> size=91, type=1
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554449
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554449: offset=33554449, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  # 33554432 is the max log file size. activemq read a location at the end of 
> the file and the size is 0, this causes the dead loop
> I don't know if this is a problem when the file is saved or there is 
> something need to be fixed when reading.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (AMQ-6936) Dead loop on log file reading

2018-03-23 Thread Gary Tully (JIRA) 

[ 
https://issues.apache.org/jira/browse/AMQ-6936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16411134#comment-16411134
 ] 

Gary Tully commented on AMQ-6936:
-

That was resolved via: 
https://git-wip-us.apache.org/repos/asf?p=activemq.git;a=blobdiff;f=activemq-kahadb-store/src/main/java/org/apache/activemq/store/kahadb/disk/journal/Journal.java;h=182a3d7757541627d67b8bf4f539836dc259bc0b;hp=da0d5b432d2eb0783b7eee686a2cb1c087124f85;hb=62bdbb0;hpb=3c342ffce43546667f59bb16db373061724e1b0f

https://issues.apache.org/jira/browse/AMQ-5603


> Dead loop on log file reading
> -
>
> Key: AMQ-6936
> URL: https://issues.apache.org/jira/browse/AMQ-6936
> Project: ActiveMQ
>  Issue Type: Bug
>  Components: KahaDB
>Affects Versions: 5.13.4
>Reporter: Rural Hunter
>Priority: Critical
>
> We restarted our activemq instance(5.13.4) but found it hung on recovering 
> data. I got the data and ran a debug and found these:
>  # The dead loop is in Journal.getNextLocation(Location location)
>  # I added log in the method and this is I got:
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554239
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554330
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554330: offset=33554330, 
> size=28, type=2
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554358: offset=33554358, 
> size=91, type=1
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554449
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554449: offset=33554449, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  # 33554432 is the max log file size. activemq read a location at the end of 
> the file and the size is 0, this causes the dead loop
> I don't know if this is a problem when the file is saved or there is 
> something need to be fixed when reading.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (AMQ-6936) Dead loop on log file reading

2018-03-23 Thread Rural Hunter (JIRA) 

[ 
https://issues.apache.org/jira/browse/AMQ-6936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16411049#comment-16411049
 ] 

Rural Hunter commented on AMQ-6936:
---

I'm not sure if it is allowed that the file size is greater than the 
maxFileLength. If it is allowed, then this might should be changed:

   else if (cur.getType() == 0) {
    // eof - jump to next datafile
    cur.setOffset(maxFileLength); => change to: 
cur.setOffset(dataFile.getLength());
    }

With the change above, my activemq instance can start up.

> Dead loop on log file reading
> -
>
> Key: AMQ-6936
> URL: https://issues.apache.org/jira/browse/AMQ-6936
> Project: ActiveMQ
>  Issue Type: Bug
>  Components: KahaDB
>Affects Versions: 5.13.4
>Reporter: Rural Hunter
>Priority: Critical
>
> We restarted our activemq instance(5.13.4) but found it hung on recovering 
> data. I got the data and ran a debug and found these:
>  # The dead loop is in Journal.getNextLocation(Location location)
>  # I added log in the method and this is I got:
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554239
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554330
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554330: offset=33554330, 
> size=28, type=2
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554358: offset=33554358, 
> size=91, type=1
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554449
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554449: offset=33554449, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  # 33554432 is the max log file size. activemq read a location at the end of 
> the file and the size is 0, this causes the dead loop
> I don't know if this is a problem when the file is saved or there is 
> something need to be fixed when reading.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

[jira] [Commented] (AMQ-6936) Dead loop on log file reading

2018-03-23 Thread Rural Hunter (JIRA) 

[ 
https://issues.apache.org/jira/browse/AMQ-6936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16411023#comment-16411023
 ] 

Rural Hunter commented on AMQ-6936:
---

These are the data files:

-rw-rw-r-- 1 activemq activemq 33554432  3月 23 11:46 db-13.log
-rw-rw-r-- 1 activemq activemq 33554432  3月 23 11:46 db-1.log
-rw-rw-r-- 1 activemq activemq 33554432  3月 23 11:46 db-2.log
-rw-rw-r-- 1 activemq activemq 33554432  3月 23 11:46 db-4.log
-rw-rw-r-- 1 activemq activemq 33554432  3月 23 11:46 db-5.log
-rw-rw-r-- 1 activemq activemq 33554454  3月 23 11:46 db-7.log
-rw-rw-r-- 1 activemq activemq    32768  3月 23 12:18 db.data
-rw-rw-r-- 1 activemq activemq    28720  3月 23 11:46 db.redo

 

db-7.log is larger than the maxFileLength = DEFAULT_MAX_FILE_LENGTH=33554432

 

> Dead loop on log file reading
> -
>
> Key: AMQ-6936
> URL: https://issues.apache.org/jira/browse/AMQ-6936
> Project: ActiveMQ
>  Issue Type: Bug
>  Components: KahaDB
>Affects Versions: 5.13.4
>Reporter: Rural Hunter
>Priority: Critical
>
> We restarted our activemq instance(5.13.4) but found it hung on recovering 
> data. I got the data and ran a debug and found these:
>  # The dead loop is in Journal.getNextLocation(Location location)
>  # I added log in the method and this is I got:
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554239
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554330
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554330: offset=33554330, 
> size=28, type=2
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554358: offset=33554358, 
> size=91, type=1
>  2018-03-23 14:56:43 [DEBUG] Journal Get next location for: 7:33554358
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554449
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554449: offset=33554449, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  2018-03-23 14:56:43 [DEBUG] Journal Reading location: 7:33554432
>  2018-03-23 14:56:43 [DEBUG] Journal Location 7:33554432: offset=33554432, 
> size=0, type=0
>  # 33554432 is the max log file size. activemq read a location at the end of 
> the file and the size is 0, this causes the dead loop
> I don't know if this is a problem when the file is saved or there is 
> something need to be fixed when reading.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)