Dear colleagues,
today I've found that one of two disks had been removed from a couple of
gmirror'ed partiotions dur to read timeout. I have done `gmirror forget'
and
`gmirror insert', and it works for relatively small partitions; however, on
rather large partiotion it did sync, and then stalls at very strage state:
root@office:/usr/local/etc# gmirror status m0g
Name Status Components
mirror/m0g DEGRADED ad4g
ad6g (100%)
root@office:/usr/local/etc# gmirror list m0g
Geom name: m0g
State: DEGRADED
Components: 2
Balance: load
Slice: 4096
Flags: NONE
GenID: 2
SyncID: 5
ID: 2249072617
Providers:
1. Name: mirror/m0g
Mediasize: 225734095872 (210G)
Sectorsize: 512
Mode: r2w1e1
Consumers:
1. Name: ad4g
Mediasize: 225734096384 (210G)
Sectorsize: 512
Mode: r1w1e1
State: ACTIVE
Priority: 0
Flags: BROKEN
GenID: 2
SyncID: 5
ID: 2403640489
2. Name: ad6g
Mediasize: 225734096384 (210G)
Sectorsize: 512
Mode: r1w1e1
State: SYNCHRONIZING
Priority: 0
Flags: DIRTY, SYNCHRONIZING
GenID: 2
SyncID: 5
Synchronized: 100%
ID: 2137028155
There is no activity on ad6g.
Any hints?
Sincerely,
D.Marck [DM5020, MCK-RIPE, DM3-RIPN]
------------------------------------------------------------------------
*** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru ***
------------------------------------------------------------------------
Hello Dmitry, You wrote on Sunday, February 18, 2007, 9:53:14 PM: DM> 1. Name: ad4g DM> Mediasize: 225734096384 (210G) DM> Sectorsize: 512 DM> Mode: r1w1e1 DM> State: ACTIVE DM> Priority: 0 DM> Flags: BROKEN Is any related messages about ad4g in /var/log/messages? -- Anton Yuzhaninov.
Dmitry Morozovsky <marck@rinet.ru> wrote:
> root@office:/usr/local/etc# gmirror status m0g
> Name Status Components
> mirror/m0g DEGRADED ad4g
> ad6g (100%)
It seems that the second disk is broken and locks up the
channel near the end of the disk.
You could try to remove the drive from the mirror and
run dd(1) on it. If the dd command also hangs near the
end of the disk, then you should definitely get a new
disk drive.
If the dd command finishes without errors (and displays
the correct number of transfered blocks), then I guess
there must be a bug in gmirror.
Best regards
Oliver
--
Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M.
Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung:
secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n-
chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart
Any opinions expressed in this message are personal to the author and may
not necessarily reflect the opinions of secnetix GmbH & Co KG in any way.
FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd
"One of the main causes of the fall of the Roman Empire was that,
lacking zero, they had no way to indicate successful termination
of their C programs."
-- Robert Firth
On Mon, 19 Feb 2007, Anton Yuzhaninov wrote: AY> DM> 1. Name: ad4g AY> DM> Mediasize: 225734096384 (210G) AY> DM> Sectorsize: 512 AY> DM> Mode: r1w1e1 AY> DM> State: ACTIVE AY> DM> Priority: 0 AY> DM> Flags: BROKEN AY> AY> Is any related messages about ad4g in /var/log/messages? There are several "TIMEOUT - READ_DMA retrying" messages, yes. However: how can gmirror continue to work with one BROKEN and other DIRTY/SYNC component? Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------