From: Lars Ellenberg <lars.ellenberg@linbit.com>
To: drbd-dev@lists.linbit.com
Cc: digimer@alteeve.com
Subject: Re: [Drbd-dev] [PATCH] scripts: Updated rhcs_fence script
Date: Wed, 13 Mar 2013 20:54:34 +0100 [thread overview]
Message-ID: <20130313195434.GF19508@soda.linbit> (raw)
In-Reply-To: <1363183325-20384-1-git-send-email-holer@ics.muni.cz>
On Wed, Mar 13, 2013 at 03:02:05PM +0100, Vlastimil Holer wrote:
> cman_tool option -f is obsolete in current RHEL 6.x.
> Syntax for kill action allows only:
> -n <nodename> The name of the node to kill (can specify multiple times)
>
> Although node is successfully fenced, rhcs_fence ends with failure
> due to wrong $exit condition handling. Part of debug log:
> rhcs_fence: 253; DEBUG: fence node2.localdomain dev 0.0 agent fence_ipmilan result: success
> fence_node[28511]: fence node2.localdomain success
> rhcs_fence: 253; DEBUG: fence node2.localdomain success
> rhcs_fence: 267; DEBUG: Attempt to fence node: [node2.localdomain] exited with: [0]
> rhcs_fence: 272; Attempt to fence: [node2.localdomain] failed!
> kernel: d-con drbd0: helper command: /sbin/drbdadm fence-peer drbd0 exit code 1 (0x100)
> kernel: d-con drbd0: fence-peer helper broken, returned 1
> ---
> scripts/rhcs_fence | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/scripts/rhcs_fence b/scripts/rhcs_fence
> index ca53678..1294660 100644
> --- a/scripts/rhcs_fence
> +++ b/scripts/rhcs_fence
> @@ -1,8 +1,8 @@
> #!/usr/bin/perl
> #
> # Author: Digimer (digimer@alteeve.com)
> -# Version: 0.2.4
> -# Released: 2012-01-20
> +# Version: 0.2.5
> +# Released: 2013-03-13
Nope, see upstream of that script:
https://github.com/digimer/rhcs_fence/commits/master
> # License: GPL v2+
> #
> # This program ties Linbit's DRBD into Red Hat's RHCS's fence daemon via the
> @@ -266,7 +266,7 @@ sub kill_target
> if ($conf->{sys}{debug}) { to_log($conf, 0, __LINE__, "DEBUG: Attempt to fence node: [$remote_node] exited with: [$sc_exit]"); }
>
> # Exit.
> - if (not $exit)
> + if ($exit)
> {
> if ($conf->{sys}{debug}) { to_log($conf, 1, __LINE__, "Attempt to fence: [$remote_node] failed!"); }
> }
That part has been fixed by the original Author already,
in a better way. We just did not notice, and did not update the
rhcs_fence as shipped with the DRBD source and package.
But that one:
> @@ -300,7 +300,7 @@ sub eject_target
> # next.
> if ($conf->{sys}{debug}) { to_log($conf, 0, __LINE__, "Target node: [$remote_node] is a cluster member, attempting to eject."); }
> $sc=IO::Handle->new();
> - $shell_call="$conf->{path}{cman_tool} kill -f $remote_node";
> + $shell_call="$conf->{path}{cman_tool} kill -n $remote_node";
digimer, what do you say?
> if ($conf->{sys}{debug}) { to_log($conf, 0, __LINE__, "DEBUG: shell call: [$shell_call]"); }
> open ($sc, "$shell_call 2>&1 |") or to_log($conf, 1, __LINE__, "Failed to call: [$sc], error was: $!");
> while(<$sc>)
> --
> 1.8.1.4
Thanks,
Lars
next prev parent reply other threads:[~2013-03-13 19:54 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-13 14:02 [Drbd-dev] [PATCH] scripts: Updated rhcs_fence script Vlastimil Holer
2013-03-13 19:54 ` Lars Ellenberg [this message]
2013-03-13 23:05 ` Madison Kelly
2013-03-15 9:26 ` Lars Ellenberg
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20130313195434.GF19508@soda.linbit \
--to=lars.ellenberg@linbit.com \
--cc=digimer@alteeve.com \
--cc=drbd-dev@lists.linbit.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox