couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Paul Joseph Davis (JIRA)" <j...@apache.org>
Subject [jira] Updated: (COUCHDB-616) 0.9->0.10 upgrade rebuilds views rather than just updating headers
Date Sat, 09 Oct 2010 19:44:50 GMT

     [ https://issues.apache.org/jira/browse/COUCHDB-616?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Paul Joseph Davis updated COUCHDB-616:
--------------------------------------

    Skill Level: Regular Contributors Level (Easy to Medium)

> 0.9->0.10 upgrade rebuilds views rather than just updating headers
> ------------------------------------------------------------------
>
>                 Key: COUCHDB-616
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-616
>             Project: CouchDB
>          Issue Type: Bug
>          Components: Database Core
>    Affects Versions: 0.10.1
>         Environment: Linux melkjug-xen 2.6.30.1-pv #1 SMP Thu Jul 9 08:12:56 EDT 2009
x86_64 Intel(R) Xeon(R) CPU E5430 @ 2.66GHz GenuineIntel GNU/Linux
> Erlang (BEAM) emulator version 5.6.5 [source] [64-bit] [smp:2] [async-threads:0] [hipe]
>            Reporter: Joshua Bronson
>
> It looks like there is code that is supposed to upgrade the headers of 0.9 views to 0.10
style when a 0.10 couch reads a 0.9 view to avoid having to rebuild them (see http://github.com/halorgium/couchdb/blob/v0.10.1/src/couchdb/couch_view_group.erl#L379).
However, they are indeed getting rebuilt.
> I have a test environment set up where I can trigger this behavior. I've added the debug
output at http://paste.pocoo.org/show/163023/ and here is what's getting printed:
> before upgrade_old_header
> after upgrade_old_header
> Pos: 0
> Error: {'EXIT',{{badmatch,eof},
>                 [{couch_file,load_header,2},
>                  {couch_file,find_header,2},
>                  {couch_file,handle_call,3},
>                  {gen_server,handle_msg,5},
>                  {proc_lib,init_p_do_apply,3}]}}
> read_header: no_valid_header
> calling reset_file
> truncating
> For some reason the file:position(Fd, eof) call is return 0, indicating a 0-length file,
but the view file in question is 188M.
> I spent a while debugging this with Jan in #couchdb today but we couldn't find the fix.

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


Mime
View raw message