You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Brandon Williams (JIRA)" <ji...@apache.org> on 2010/04/29 23:18:54 UTC

[jira] Created: (CASSANDRA-1038) Not all column families are created

Not all column families are created
-----------------------------------

                 Key: CASSANDRA-1038
                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1038
             Project: Cassandra
          Issue Type: Bug
          Components: Core
    Affects Versions: 0.7
            Reporter: Brandon Williams
             Fix For: 0.7


It seems that not all column families will be created via system_add_keyspace in some cases.  To reproduce:

Run stress.py (with CASSANDRA-1033) inserts (I used 1M) against standard columns.  During this run, both Standard1 and Super1 will be created.

Run stress.py again, this time against super columns.  Due to CASSANDRA-1036 no errors will be visible to the client but can be observed in the log.

You can switch the order and stress supers first, in which case Standard1 will not exist.  If you call describe_keyspace on Keyspace1, it will show both CFs even though only one will work.


-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Assigned: (CASSANDRA-1038) Not all column families are created

Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/CASSANDRA-1038?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jonathan Ellis reassigned CASSANDRA-1038:
-----------------------------------------

    Assignee: Gary Dusbabek

> Not all column families are created
> -----------------------------------
>
>                 Key: CASSANDRA-1038
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1038
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.7
>            Reporter: Brandon Williams
>            Assignee: Gary Dusbabek
>             Fix For: 0.7
>
>
> It seems that not all column families will be created via system_add_keyspace in some cases.  To reproduce:
> Run stress.py (with CASSANDRA-1033) inserts (I used 1M) against standard columns.  During this run, both Standard1 and Super1 will be created.
> Run stress.py again, this time against super columns.  Due to CASSANDRA-1036 no errors will be visible to the client but can be observed in the log.
> You can switch the order and stress supers first, in which case Standard1 will not exist.  If you call describe_keyspace on Keyspace1, it will show both CFs even though only one will work.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Updated: (CASSANDRA-1038) Not all column families are created

Posted by "Gary Dusbabek (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/CASSANDRA-1038?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Gary Dusbabek updated CASSANDRA-1038:
-------------------------------------

    Attachment: 0001-restrict-how-and-when-CFMetaData-objects-are-added-t.patch

The problem was that when the keyspace creation was reattempted, identical CFMetaData objects were created that obliterated the older, correct cfids.

I could have put a check in the private CFMetaData constructor that threw an exception, but decided I'd rather have explicit control over cfIdMap instead.

> Not all column families are created
> -----------------------------------
>
>                 Key: CASSANDRA-1038
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1038
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.7
>            Reporter: Brandon Williams
>            Assignee: Gary Dusbabek
>             Fix For: 0.7
>
>         Attachments: 0001-restrict-how-and-when-CFMetaData-objects-are-added-t.patch
>
>
> It seems that not all column families will be created via system_add_keyspace in some cases.  To reproduce:
> Run stress.py (with CASSANDRA-1033) inserts (I used 1M) against standard columns.  During this run, both Standard1 and Super1 will be created.
> Run stress.py again, this time against super columns.  Due to CASSANDRA-1036 no errors will be visible to the client but can be observed in the log.
> You can switch the order and stress supers first, in which case Standard1 will not exist.  If you call describe_keyspace on Keyspace1, it will show both CFs even though only one will work.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Commented: (CASSANDRA-1038) Not all column families are created

Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/CASSANDRA-1038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12864358#action_12864358 ] 

Jonathan Ellis commented on CASSANDRA-1038:
-------------------------------------------

+1

> Not all column families are created
> -----------------------------------
>
>                 Key: CASSANDRA-1038
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1038
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.7
>            Reporter: Brandon Williams
>            Assignee: Gary Dusbabek
>             Fix For: 0.7
>
>         Attachments: 0001-restrict-how-and-when-CFMetaData-objects-are-added-t.patch
>
>
> It seems that not all column families will be created via system_add_keyspace in some cases.  To reproduce:
> Run stress.py (with CASSANDRA-1033) inserts (I used 1M) against standard columns.  During this run, both Standard1 and Super1 will be created.
> Run stress.py again, this time against super columns.  Due to CASSANDRA-1036 no errors will be visible to the client but can be observed in the log.
> You can switch the order and stress supers first, in which case Standard1 will not exist.  If you call describe_keyspace on Keyspace1, it will show both CFs even though only one will work.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.