[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-users] Xen Configuration questions
Hello, Happy New Year everybody. I've been coding during much of the winter break, whilst stuck indoors with bronchitis, but I'm now trying to catch up on my mailbox a bit! > >> 1). Is RHEL & Xen ready for prime time? (is it production ready, will > >> I sleep at night?) Well, RedHat seem to think so ;-) It rather depends what you're doing though. For instance, Windows guests are only supported since RHEL 5.1 - and those won't perform fantastically unless you get the driver pack that Redhat are going to (or have already?) release. I'm not sure if they officially support Live Migration or some of the other whizzy features. > >> 2). Each Xen instance will have a LUN on the iSCSI. Should I connect > >> the LUN on Domain0 and then have the DomU load it or should I have > >> each DomU run iSCSI directly with the SAN? I want to enable live > >> migrate between the blades. I always think it's nicer to abstract away details of storage from guests as much as possible and handle them all in dom0 - that way domU just sees there's a "VBD" and doesn't need to know if you change things. It also avoids having to do funny things to make iSCSI root work. There are probably good arguments for exposing iSCSI to the guests too, so it depends on your preference, I guess. > >> 3). Are there any pre-built Xen 'Appliances' for common things like > >> DNS resolvers? rpath distribute a whole load of appliances, with options for Xen support as well as support for other VMMs and native install. These are fairly polished: download image, boot guest, DHCP automatically, prompted for configuration info, further config via web interface. jailtime.org and xen-get.org both have some images but some of these may well be a bit out of date. > >> 4). Should I use rsync between the blades to syn the xen configs to I > >> can live migrate guest machines between the blades? > >> > >> My primary concern is #1, if it isn't stable I'll look to VMWare or > >> Virtuozzo. I have a machine running Virtuozzo already and it works > >> ok, Would rather run something open though, Linux has served me well > >> the last 17 years. If you want something open and you like Virtuozze, there's always OpenVZ? Also, you've got the option of various other virtualisers (KVM, the free-beer VMware Server) at no cost, which you might also find suitable. Finally, there's a range of different Xen-based hypervisor appliances of varying levels of open / closed coming onto the market from Novell, XenSource/Citrix, Virtual Iron, Oracle, Sun, etc. > I've been using RHEL 5 for Xen servers in production. I don't > live-migrate right now because I'm not using (or rather, didn't use in > my last contract) network based Xen images, but I did keep the > configurations distributed on all the servers for load balancing and > restoration from backup. Even doing non-live migration - or shutdown on one machine, then restart on another - should help make life easier. > I'm not aware of Xen appliances, but for DNS servers you can use a > stripped down RHEL with bind and nscd pretty trivially. Stripping an > RHEL system to bare bones is a pain due to the dependency hel and > massive overpopulation of the "Basic" installation. Frankly, a > bare-bones install doesn't even need Perl, but try teaching the RPM > dependency tree not do that. And try telling an RHEL installer that no, > you don't *WANT* NIS and PCMCIA and Bluetooth and a bunch of other silly > things in a Xen environment. I wonder if anybody has put together a script to "barebonesify" RHEL / CentOS - that would be quite useful! Cheers, Mark -- Dave: Just a question. What use is a unicyle with no seat? And no pedals! Mark: To answer a question with a question: What use is a skateboard? Dave: Skateboards have wheels. Mark: My wheel has a wheel! _______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-users
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |