000940.html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
 <HEAD>
   <TITLE> [zfs-discuss] chksum error on RAID.Z - severity?
   </TITLE>
   <LINK REL="Index" HREF="index.html" >
   <LINK REL="made" HREF="mailto:zfs-discuss%40lists.macosforge.org?Subject=%5Bzfs-discuss%5D%20chksum%20error%20on%20RAID.Z%20-%20severity%3F&In-Reply-To=DB9A2F12-AD43-41B9-913A-23DA8977D519%40hessmann.de">
   <META NAME="robots" CONTENT="index,nofollow">
   <META http-equiv="Content-Type" content="text/html; charset=us-ascii">
   <LINK REL="Previous"  HREF="000939.html">
   <LINK REL="Next"  HREF="000941.html">
 </HEAD>
 <BODY BGCOLOR="#ffffff">
   <H1>[zfs-discuss] chksum error on RAID.Z - severity?</H1>
    <B>Christian He&#223;mann</B> 
    <A HREF="mailto:zfs-discuss%40lists.macosforge.org?Subject=%5Bzfs-discuss%5D%20chksum%20error%20on%20RAID.Z%20-%20severity%3F&In-Reply-To=DB9A2F12-AD43-41B9-913A-23DA8977D519%40hessmann.de"
       TITLE="[zfs-discuss] chksum error on RAID.Z - severity?">zfs at hessmann.de
       </A><BR>
    <I>Tue Oct  7 06:27:25 PDT 2008</I>
    <P><UL>
        <LI>Previous message: <A HREF="000939.html">[zfs-discuss] Filesystem operations hang
</A></li>
        <LI>Next message: <A HREF="000941.html">[zfs-discuss] chksum error on RAID.Z - severity?
</A></li>
         <LI> <B>Messages sorted by:</B> 
              <a href="date.html#940">[ date ]</a>
              <a href="thread.html#940">[ thread ]</a>
              <a href="subject.html#940">[ subject ]</a>
              <a href="author.html#940">[ author ]</a>
         </LI>
       </UL>
    <HR>  
<!--beginarticle-->
<PRE>On 26.09.2008, at 10:26, Christian He&#223;mann wrote:

&gt; Me again...
&gt;
&gt; After adding three more disks to my RAID.Z Pool (now 2x 3 disks) I
&gt; issued a scrub just to be sure everything is fine.
&gt;
&gt; Well, apparently it isn't, since zpool tells me I have chksum errors
&gt; (85% done, 5 errors so far) on one of the additional disks (which
&gt; isn't new and the smallest of all: old 3x 750GB, new 500/200/120GB).
&gt; That same disk worked fine as a standalone ZFS device until yesterday,
&gt; with scrubs without any error messages.

Well, seems I have to talk to myself, since no one wants to answer 
me. ;-)

I decided not to replace the disk and didn't encounter any further  
problems (didn't do another scrub, though, due to the bug which causes  
the scrub to restart as soon as a snapshot is added - any chance for  
its fix to be implemented in OS X soon?).

After a planned reboot and therefore new mount, I issued another  
scrub, with a nearly completely filled pool (~5GB available) and: no  
more chksum errors.

I don't believe the blocks in question were all free during that time,  
so either the disk marked them as bad (don't think so) or it was not a  
sign for a bad disk, but some other problem - maybe during write/read,  
maybe due to the kernel panics the system had to endure, maybe one of  
the ZFS bugs displaying chksum errors which aren't there...

Anyway, if anyone is interested, my personal conclusion: Don't Panic!

Best regards,

Christian
</PRE>

<!--endarticle-->
    <HR>
    <P><UL>
        <!--threads-->
	<LI>Previous message: <A HREF="000939.html">[zfs-discuss] Filesystem operations hang
</A></li>
	<LI>Next message: <A HREF="000941.html">[zfs-discuss] chksum error on RAID.Z - severity?
</A></li>
         <LI> <B>Messages sorted by:</B> 
              <a href="date.html#940">[ date ]</a>
              <a href="thread.html#940">[ thread ]</a>
              <a href="subject.html#940">[ subject ]</a>
              <a href="author.html#940">[ author ]</a>
         </LI>
       </UL>

<hr>
<a href="http://lists.macosforge.org/mailman/listinfo.cgi/zfs-discuss">More information about the zfs-discuss
mailing list</a><br>
</body></html>

/__MAILINGLIST/lists.macosforge.org/pipermail/zfs-discuss/2008-October/000940.html