Hi I found the answer in a manual, of course. I should have looked in these first... sorry.
http://publib.boulder.ibm.com/infocenter/zvm/v6r1/index.jsp search for 'spanning tree' •How does this affect the Virtual Switch? The virtual switch (VSWITCH) does not participate in the Spanning Tree Protocol. The VSWITCH can be thought of as an edge switch. The VSWITCH connects the z/VM host LAN segment to the external network. This configuration typically does not have redundant paths from each host. Due to the nature of z/VM virtual networks, redundant paths (loops) can be easily remedied through configuration changes. Avoiding dual path situations caused by routing hosts forwarding broadcasts in a virtual switched network is the key to preventing broadcast storms. Redundancy is rarely required because recovery of network outages on OSA-Express ports is transparently provided by the VSWITCH. Some suggestions: ◦Problems can arise if a guest on the VSWITCH has a separate connection to the external network on the same subnet. To avoid this, guest's connections to the external network should be through the VSWITCH only. If a need arises to keep some traffic isolated from other guests on the virtual switch, VLANs should be used. ◦Configuration problems with a link aggregation group on the physical switch could also lead to problems if spanning tree protocol is not in place on the physical switch. Symptoms would include forwarding multiple copies of packets into the VSWITCH network. However, the virtual system implementation avoids aggravating the problem eliminating loop back from the link aggregation port. Packets that have been received on the link aggregation port are not forwarded back onto the physical network. ◦When a z/VM host is acting as a router over to separate VSWITCHs (two separate LAN segments), you should ensure that the router is not configured to forward broadcasts from one LAN segment to the other. BR /Tore ________________________________________________ Tore Agblad System programmer, Volvo IT certified IT Architect Volvo Information Technology Infrastructure Mainframe Design & Development, Linux servers Dept 4253 DA1S SE-405 08, Gothenburg Sweden Telephone: +46-31-3233569 E-mail: [email protected] http://www.volvo.com/volvoit/global/en-gb/ -----Original Message----- From: Agblad Tore Sent: den 15 maj 2012 11:28 To: 'Linux on 390 Port' Subject: Anybody know if VSWITCHes support 'spanning tree' , or need to do it ? We have that question from our network people, because we will need to connect to a separate physical LAN for one customer, and we want the migration as seamless as possible. So we will have duplicates of connections, of which both are redundant as well. This will obviously expose us for the risk of network routing loops, and this is prohibited with 'spanning tree' functionality, normally supported by switches or routers. Any know if a z/VM vswitch support this or if it is perhaps not needed ? Anyone having experinces of this ? We run z/VM 6.1 and SLES11.1 and RHEL6 BR /Tore ________________________________________________ Tore Agblad System programmer, Volvo IT certified IT Architect Volvo Information Technology Infrastructure Mainframe Design & Development, Linux servers Dept 4253 DA1S SE-405 08, Gothenburg Sweden Telephone: +46-31-3233569 E-mail: [email protected] http://www.volvo.com/volvoit/global/en-gb/ -----Original Message----- From: Linux on 390 Port [mailto:[email protected]] On Behalf Of Alan Altmark Sent: den 10 maj 2012 16:03 To: [email protected] Subject: Re: SLES 11 SP1 msg: kernel: scsi: host 0 channel 0 id 0 lun2 has a LUN larger than allowed by the host adapter On Wednesday, 05/09/2012 at 09:39 EDT, Raymond Higgs/Poughkeepsie/IBM@IBMUS wrote: > > 11:12:11 scsi: host 0 channel 0 id 0 lun2 has a LUN larger than > > allowed by the host adapter > There shouldn't be any limitations until 2 TB. I don't think linux is > complaining about the size of the lun. I think linux is complaining about > the lun number. It is saying that it will only configure luns 0x0000 and > 0x0001. Nominally, that response means that the LUN number of one of the LUNs returned on REPORT LUNS is larger than the value supported by the adapter, which zfcp_scsi_adapter_register() sets to 0xFFFFFFFF. Those are (supposed to be) unsigned ints, so the implication is, as you suggested, that something changed this value via sysfs, assuming it is exported. From drivers/scsi/scsi_scan.c: 1466 if (lun > sdev->host->max_lun) { 1467 printk(KERN_WARNING "scsi: %s lun%d has a LUN larger" 1468 " than allowed by the host adapter\n", 1469 devname, lun); Alan Altmark Senior Managing z/VM and Linux Consultant IBM System Lab Services and Training ibm.com/systems/services/labservices office: 607.429.3323 mobile; 607.321.7556 [email protected] IBM Endicott ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
