<font size=2 face="Verdana">Hi Alexander,</font>
<br>
<br><font size=2 face="Verdana">thanks for replying, here's my current
multipath.conf:</font>
<br>
<br><font size=2 face="Verdana">defaults {</font>
<br><font size=2 face="Verdana"> user_friendly_names
yes</font>
<br><font size=2 face="Verdana">}</font>
<br>
<br><font size=2 face="Verdana">blacklist {</font>
<br><font size=2 face="Verdana"> devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"</font>
<br><font size=2 face="Verdana"> devnode "^(hd|xvd|vd)[a-z]*"</font>
<br><font size=2 face="Verdana"> wwid "*"</font>
<br><font size=2 face="Verdana">}</font>
<br>
<br><font size=2 face="Verdana">blacklist_exceptions {</font>
<br><font size=2 face="Verdana"> wwid "3600144f0fdf58b5c00004bc738070001"</font>
<br>
<br><font size=2 face="Verdana"> devices{</font>
<br><font size=2 face="Verdana">
device {</font>
<br><font size=2 face="Verdana">
vendor "SUN"</font>
<br><font size=2 face="Verdana">
product "Sun Storage 7310"</font>
<br><font size=2 face="Verdana">
}</font>
<br><font size=2 face="Verdana"> }</font>
<br><font size=2 face="Verdana">}</font>
<br>
<br><font size=2 face="Verdana">devices {</font>
<br><font size=2 face="Verdana"> device
{</font>
<br><font size=2 face="Verdana">
vendor "SUN"</font>
<br><font size=2 face="Verdana">
product "Sun Storage 7310"</font>
<br><font size=2 face="Verdana">
path_grouping_policy failover</font>
<br><font size=2 face="Verdana">
getuid_callout "/sbin/scsi_id -g -u -s /block/%n"</font>
<br><font size=2 face="Verdana">
prio_callout "/sbin/mpath_prio_rdac /dev/%n"</font>
<br><font size=2 face="Verdana">
features "0"</font>
<br><font size=2 face="Verdana">
failback immediate</font>
<br><font size=2 face="Verdana">
rr_weight uniform</font>
<br><font size=2 face="Verdana">
no_path_retry queue</font>
<br><font size=2 face="Verdana">
rr_min_io 1000</font>
<br><font size=2 face="Verdana"> }</font>
<br><font size=2 face="Verdana">}</font>
<br>
<br><font size=2 face="Verdana">I added "path_grouping_policy failover"
because of your message. I also noticed, you have path_grouping_policy
specified twice; is this on purpose?</font>
<br><font size=2 face="Verdana">Also, when I activate </font><tt><font size=2>'hardware_handler
"1 rdac"</font></tt><font size=2 face="Verdana">', the box does
not boot any more with some rdac driver error message that I can catch
since it scrolls by fast...</font>
<br>
<br><font size=2 face="Verdana">with the above multipath.conf it gets even
stranger:</font>
<br>
<br><font size=2 face="Verdana">[root@dev-db1 ~]# multipath -ll</font>
<br><font size=2 face="Verdana">sdd: checker msg is "readsector0 checker
reports path is down"</font>
<br><font size=2 face="Verdana">sdh: checker msg is "readsector0 checker
reports path is down"</font>
<br><font size=2 face="Verdana">sdi: checker msg is "readsector0 checker
reports path is down"</font>
<br><font size=2 face="Verdana">sdj: checker msg is "readsector0 checker
reports path is down"</font>
<br><font size=2 face="Verdana">C9 Inquiry of device </dev/sda> failed.</font>
<br><font size=2 face="Verdana">C9 Inquiry of device </dev/sde> failed.</font>
<br><font size=2 face="Verdana">mpath0 (3600144f0fdf58b5c00004bc738070001)
dm-0 SUN,Sun Storage 7310</font>
<br><font size=2 face="Verdana">[size=50G][features=1 queue_if_no_path][hwhandler=0][rw]</font>
<br><font size=2 face="Verdana">\_ round-robin 0 [prio=0][active]</font>
<br><font size=2 face="Verdana"> \_ 0:0:1:0 sda 8:0 [active][ready]
</font>
<br><font size=2 face="Verdana"> \_ 1:0:0:0 sde 8:64 [active][ready]</font>
<br>
<br><font size=2 face="Verdana">however, the system "lives happily",
despite the error messages.</font>
<br>
<br><font size=2 face="Verdana">Jens Neu<br>
Health Services Network Administration<br>
<br>
Phone: +49 (0) 30 68905-2412<br>
Mail: jens.neu@biotronik.de</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="Verdana"><b>"Alexander Dalloz"
<ad+lists@uni-x.org></b> </font>
<br><font size=1 face="Verdana">Sent by: centos-bounces@centos.org</font>
<p><font size=1 face="Verdana">05/27/2010 04:47 PM</font>
<table border>
<tr valign=top>
<td bgcolor=white>
<div align=center><font size=1 face="Verdana">Please respond to<br>
CentOS mailing list <centos@centos.org></font></div></table>
<br>
<td width=59%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="Verdana">To</font></div>
<td><font size=1 face="Verdana">"CentOS mailing list" <centos@centos.org></font>
<tr valign=top>
<td>
<div align=right><font size=1 face="Verdana">cc</font></div>
<td>
<tr valign=top>
<td>
<div align=right><font size=1 face="Verdana">Subject</font></div>
<td><font size=1 face="Verdana">Re: [CentOS] Multipathing with Sun 7310</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><tt><font size=2>> Dear list,<br>
><br>
> we have a relatively new Sun Storage 7310, where we connect CentOS
5.5<br>
> Servers (IBM LS21/LS41 Blades) via Brocade Switches, 4GBit FC. The
Blades<br>
> boot from SAN via qla2xxx, and have no harddisks at all. We want them
to<br>
> use multipathing from the very beginning, so /boot and / are already
seen<br>
> by multipathd. Problem is, that the Sun 7310 has two storage heads
which<br>
> run in active/passive mode. BUT the multipathd thinks, they are<br>
> active/active and therefor shows half the available paths as faulty<br>
> (multipath -ll below)<br>
> While this probably gives me the redundancy that is desired, it is
a<br>
> relatively messy situation, since it will be unnecessary hard to detect<br>
> real path failures and the OS is complaining about "readsector0
checker<br>
> reports path is down" which gives me >40M/24h /var/log/messages
garbage.<br>
> Any hints for a reasonable configuration? Unfortunately the Sun 7310
is<br>
> rather new, so almost nothing shows up on google... even less for<br>
> RHEL/CentOS :-(<br>
><br>
> regards from Berlin<br>
> Jens<br>
><br>
> [root@dev-db1 tmp]# multipath -ll<br>
> sdaa: checker msg is "readsector0 checker reports path is down"<br>
> sdab: checker msg is "readsector0 checker reports path is down"<br>
> sdac: checker msg is "readsector0 checker reports path is down"<br>
> sdad: checker msg is "readsector0 checker reports path is down"<br>
> sdd: checker msg is "readsector0 checker reports path is down"<br>
> sdh: checker msg is "readsector0 checker reports path is down"<br>
> sdl: checker msg is "readsector0 checker reports path is down"<br>
> sdp: checker msg is "readsector0 checker reports path is down"<br>
> sdq: checker msg is "readsector0 checker reports path is down"<br>
> sdr: checker msg is "readsector0 checker reports path is down"<br>
> sds: checker msg is "readsector0 checker reports path is down"<br>
> sdt: checker msg is "readsector0 checker reports path is down"<br>
> sdu: checker msg is "readsector0 checker reports path is down"<br>
> sdv: checker msg is "readsector0 checker reports path is down"<br>
> sdx: checker msg is "readsector0 checker reports path is down"<br>
> sdz: checker msg is "readsector0 checker reports path is down"<br>
> mpath0 (3600144f0fdf58b5c00004bc738070001) dm-0 SUN,Sun Storage 7310<br>
> [size=50G][features=0][hwhandler=0][rw]<br>
> \_ round-robin 0 [prio=1][active]<br>
> \_ 0:0:1:0 sda 8:0 [active][ready]<br>
> \_ round-robin 0 [prio=1][enabled]<br>
> \_ 1:0:0:0 sde 8:64 [active][ready]<br>
> \_ round-robin 0 [prio=1][enabled]<br>
> \_ 0:0:2:0 sdi 8:128 [active][ready]<br>
> \_ round-robin 0 [prio=1][enabled]<br>
> \_ 1:0:1:0 sdm 8:192 [active][ready]<br>
> \_ round-robin 0 [prio=0][enabled]<br>
> \_ 0:0:3:0 sdq 65:0 [failed][faulty]<br>
> \_ round-robin 0 [prio=0][enabled]<br>
> \_ 1:0:2:0 sdr 65:16 [failed][faulty]<br>
> \_ round-robin 0 [prio=0][enabled]<br>
> \_ 0:0:4:0 sdx 65:112 [failed][faulty]<br>
> \_ round-robin 0 [prio=0][enabled]<br>
> \_ 1:0:3:0 sdz 65:144 [failed][faulty]<br>
<br>
Hello Jens,<br>
<br>
looks like your multipathing setup is using multibus path_grouping_policy.<br>
Did you create a custom /etc/multipath.conf? You should have. As your<br>
device is new and not known with proper defaults by the<br>
device-mapper-multipath you will have to set the path_grouping_policy<br>
explicitly.<br>
<br>
devices {<br>
device {<br>
vendor "SUN"<br>
product "Sun
Storage 7310" <== REMARK<br>
path_grouping_policy
failover<br>
getuid_callout
"/sbin/scsi_id -g -u -s /block/%n"<br>
prio_callout "/sbin/mpath_prio_rdac
/dev/%n"<br>
features "0"<br>
hardware_handler
"1 rdac"<br>
path_grouping_policy
group_by_prio<br>
failback immediate<br>
rr_weight uniform<br>
no_path_retry
queue<br>
rr_min_io 1000<br>
path_checker rdac<br>
}<br>
}<br>
<br>
REMARK: to detect the proper product name issue for instance:<br>
<br>
cat /sys/block/sda/device/model<br>
<br>
Restart the multipathd after editing the multipath.conf and I expect<br>
`multipath -ll" or `multipathd -k"show paths"' will show
you the paths<br>
correctly weighted.<br>
<br>
Regards<br>
<br>
Alexander<br>
<br>
<br>
_______________________________________________<br>
CentOS mailing list<br>
CentOS@centos.org<br>
</font></tt><a href=http://lists.centos.org/mailman/listinfo/centos><tt><font size=2>http://lists.centos.org/mailman/listinfo/centos</font></tt></a><tt><font size=2><br>
</font></tt>
<br>