From mboxrd@z Thu Jan 1 00:00:00 1970 From: christophe varoqui Subject: RE: [dm-devel] [ANNOUNCE] multipath-tools-0.4.1 Date: Wed, 22 Dec 2004 02:26:12 +0100 Message-ID: <1103678772.6030.11.camel@zezette> References: <1103677780.6030.4.camel@zezette> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <1103677780.6030.4.camel@zezette> Sender: linux-raid-owner@vger.kernel.org To: device-mapper development Cc: "linux-raid@vger.kernel.org" , "linux-hotplug-devel@lists.sourceforge.net" , "linux-scsi@vger.kernel.org" List-Id: linux-raid.ids Le mercredi 22 d?embre 2004 =C3=A0 02:09 +0100, christophe varoqui a =C3= =A9crit : > Le mardi 21 d?embre 2004 =C3=A0 14:02 -0800, Caushik, Ramesh a =C3=A9= crit : > > The devinfo.c file in the multipath-tools-0.4.1 appears to have a t= ypo > > in the code to get the node_name attr in the fc_transport sysfs ent= ry. > > Path below should fix it. BTW why does a failure to get a node_name > > attribute, result in failure of multipath discovery (because devinf= o > > returns failure)even if the "group_by_node_name" policy is not used= ? > > Can't we just NULL out the tgt_node_name string in the sysfs_devinf= o > > routine and fail the group_by_node_name routine if that policy was > > chosen ? That is what happened in my case. Multipath discovery fail= ed > > due to above reason even though node_name grouping was not specifie= d.=20 > >=20 > You are absolutely right. >=20 > I was aware that introducing that FC-ism would break for iSCSI and ot= her > transports. I just thought I'll get that right in a next release beca= use > I wasn't aware someone was using the thing in such an "exotic" > environment. >=20 > I would gladly get a description of your topology, by the way. > It would help to get it right in the future. >=20 > Anyway, I'll fix that for 0.4.2 >=20 That response was for the premature exit when the FC-ism is out of place. Concerning the patch itself, I don't see the problem it adresses : [root@cl039 multipath]# ll /sys/class/fc_transport/ total 0 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:0 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:1 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:2 drwxr-xr-x 2 root root 0 Dec 14 16:27 target2:0:3 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:0 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:1 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:2 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:3 [root@cl039 multipath]# ll /sys/class/fc_transport/target2\:0\:0 total 0 lrwxrwxrwx 1 root root 0 Dec 14 16:26 device -> ../../../devices/pci0000:00/0000:00:09.0/0000:03:04.0/host2/target2:= 0:0 -r--r--r-- 1 root root 4096 Dec 14 16:26 node_name -r--r--r-- 1 root root 4096 Dec 14 16:26 port_id -r--r--r-- 1 root root 4096 Dec 14 16:26 port_name This is with Qlogic hardware, but I'm not aware of HBA driver specifics in the area. If there is, I'm all for resolving the delta at the device driver level. > regards, --=20 christophe varoqui - To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html From mboxrd@z Thu Jan 1 00:00:00 1970 From: christophe varoqui Date: Tue, 21 Dec 2004 23:19:13 +0000 Subject: RE: [dm-devel] [ANNOUNCE] multipath-tools-0.4.1 Message-Id: <1103678772.6030.11.camel@zezette> List-Id: References: <1103677780.6030.4.camel@zezette> In-Reply-To: <1103677780.6030.4.camel@zezette> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit To: device-mapper development Cc: "linux-raid@vger.kernel.org" , "linux-hotplug-devel@lists.sourceforge.net" , "linux-scsi@vger.kernel.org" Le mercredi 22 d?embre 2004 à 02:09 +0100, christophe varoqui a écrit : > Le mardi 21 d?embre 2004 à 14:02 -0800, Caushik, Ramesh a écrit : > > The devinfo.c file in the multipath-tools-0.4.1 appears to have a typo > > in the code to get the node_name attr in the fc_transport sysfs entry. > > Path below should fix it. BTW why does a failure to get a node_name > > attribute, result in failure of multipath discovery (because devinfo > > returns failure)even if the "group_by_node_name" policy is not used ? > > Can't we just NULL out the tgt_node_name string in the sysfs_devinfo > > routine and fail the group_by_node_name routine if that policy was > > chosen ? That is what happened in my case. Multipath discovery failed > > due to above reason even though node_name grouping was not specified. > > > You are absolutely right. > > I was aware that introducing that FC-ism would break for iSCSI and other > transports. I just thought I'll get that right in a next release because > I wasn't aware someone was using the thing in such an "exotic" > environment. > > I would gladly get a description of your topology, by the way. > It would help to get it right in the future. > > Anyway, I'll fix that for 0.4.2 > That response was for the premature exit when the FC-ism is out of place. Concerning the patch itself, I don't see the problem it adresses : [root@cl039 multipath]# ll /sys/class/fc_transport/ total 0 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:0 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:1 drwxr-xr-x 2 root root 0 Dec 14 16:26 target2:0:2 drwxr-xr-x 2 root root 0 Dec 14 16:27 target2:0:3 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:0 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:1 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:2 drwxr-xr-x 2 root root 0 Dec 14 16:27 target3:0:3 [root@cl039 multipath]# ll /sys/class/fc_transport/target2\:0\:0 total 0 lrwxrwxrwx 1 root root 0 Dec 14 16:26 device -> ../../../devices/pci0000:00/0000:00:09.0/0000:03:04.0/host2/target2:0:0 -r--r--r-- 1 root root 4096 Dec 14 16:26 node_name -r--r--r-- 1 root root 4096 Dec 14 16:26 port_id -r--r--r-- 1 root root 4096 Dec 14 16:26 port_name This is with Qlogic hardware, but I'm not aware of HBA driver specifics in the area. If there is, I'm all for resolving the delta at the device driver level. > regards, -- christophe varoqui ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://productguide.itmanagersjournal.com/ _______________________________________________ Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net Linux-hotplug-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel