FAQ
I had one out of two nodes shutdown last night due to a full file
system on the server.
I've cleaned up the disk space, and restarted the node with ndbd, but
the node never finishes starting. I've given it well over an hour, it
used to take maybe 20 minutes to start. The log shows the last
message from the node on the management node (different server):
Start phase 4 completed (initial node restart)

There are no log messages in the nodes logs.

I also tried ndbd --initial ( which makes me a bit nervous but has
worked before ).

Any suggestions?

Eric

Search Discussions

  • Eric Williams at Feb 16, 2007 at 2:48 pm
    On Feb 16, 2007, at 9:08 AM, Eric Williams wrote:
    I had one out of two nodes shutdown last night due to a full file
    system on the server.
    I've cleaned up the disk space, and restarted the node with ndbd, but
    the node never finishes starting. I've given it well over an hour, it
    used to take maybe 20 minutes to start. The log shows the last
    message from the node on the management node (different server):
    Start phase 4 completed (initial node restart)

    There are no log messages in the nodes logs.

    I should say there are node new message in the logs regarding errors
    when restarting.


    I also tried ndbd --initial ( which makes me a bit nervous but has
    worked before ).

    Currently I have killed the last startup with --initial, since it was
    taking so long.
    I have restarted the node with just ndbd.


    Any suggestions?

    I suppose it is possible I'm not being patient enough? If I'm not
    seeing any negative log messages, should I just assume phase 5 is
    taking this long? That would make sense I guess, but I probably
    panicked.

    Eric
  • Eric Williams at Feb 16, 2007 at 2:54 pm

    On Feb 16, 2007, at 9:55 AM, Jonas Oreland wrote:

    Eric Williams wrote:
    On Feb 16, 2007, at 9:08 AM, Eric Williams wrote:
    I had one out of two nodes shutdown last night due to a full file
    system on the server.
    I've cleaned up the disk space, and restarted the node with ndbd,
    but the node never finishes starting. I've given it well over an
    hour, it used to take maybe 20 minutes to start. The log shows the
    last message from the node on the management node (different
    server): Start phase 4 completed (initial node restart)
    There are no log messages in the nodes logs.
    I should say there are node new message in the logs regarding
    errors when restarting.
    I also tried ndbd --initial ( which makes me a bit nervous but has
    worked before ).
    Currently I have killed the last startup with --initial, since it
    was taking so long.
    I have restarted the node with just ndbd.
    Any suggestions?
    I suppose it is possible I'm not being patient enough? If I'm not
    seeing any negative log messages, should I just assume phase 5 is
    taking this long? That would make sense I guess, but I probably
    panicked.
    Eric
    which version?

    Ugh, sorry. 5.0.27.

    Thanks

    Eric
  • Jonas Oreland at Feb 16, 2007 at 3:08 pm

    Eric Williams wrote:
    On Feb 16, 2007, at 9:55 AM, Jonas Oreland wrote:

    Eric Williams wrote:
    On Feb 16, 2007, at 9:08 AM, Eric Williams wrote:
    I had one out of two nodes shutdown last night due to a full file
    system on the server.
    I've cleaned up the disk space, and restarted the node with ndbd, but
    the node never finishes starting. I've given it well over an hour, it
    used to take maybe 20 minutes to start. The log shows the last
    message from the node on the management node (different server):
    Start phase 4 completed (initial node restart)
    There are no log messages in the nodes logs.
    I should say there are node new message in the logs regarding errors
    when restarting.
    I also tried ndbd --initial ( which makes me a bit nervous but has
    worked before ).
    Currently I have killed the last startup with --initial, since it was
    taking so long.
    I have restarted the node with just ndbd.
    Any suggestions?
    I suppose it is possible I'm not being patient enough? If I'm not
    seeing any negative log messages, should I just assume phase 5 is
    taking this long? That would make sense I guess, but I probably
    panicked.
    Eric
    which version?

    Ugh, sorry. 5.0.27.
    looks like, http://bugs.mysql.com/bug.php?id=20895

    if it is, you can locate "master" node in output from ndb_mgm -e show
    and do "X dump 7098 0 0" where X = node id of master node,

    if it isnt, this will likely crash your cluster

    /Jonas
    Thanks

    Eric


    --MySQL Cluster Mailing List
    For list archives: http://lists.mysql.com/cluster
    To unsubscribe: http://lists.mysql.com/cluster?unsub=jonas@mysql.com

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupcluster @
categoriesmysql
postedFeb 16, '07 at 2:03p
activeFeb 16, '07 at 3:08p
posts4
users2
websitemysql.com
irc#mysql

2 users in discussion

Eric Williams: 3 posts Jonas Oreland: 1 post

People

Translate

site design / logo © 2018 Grokbase