<br><br><div class="gmail_quote">On Tue, Mar 17, 2009 at 1:03 PM, brent timothy saner <span dir="ltr"><<a href="mailto:brent.saner@gmail.com">brent.saner@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
-----BEGIN PGP SIGNED MESSAGE-----<br>
Hash: SHA1<br>
<div class="im"><br>
Victor Padro wrote:<br>
> Hello,<br>
><br>
> I'm trying to setup loadbalanced high-availability apache cluster based<br>
> on ubuntu 8.04 using this how-to<br>
</div>> <<a href="http://www.howtoforge.com/set-up-a-loadbalanced-ha-apache-cluster-ubuntu8.04" target="_blank">http://www.howtoforge.com/set-up-a-loadbalanced-ha-apache-cluster-ubuntu8.04</a>>,<br>
<div class="im">> using a couple of openvz containers, but I get stuck when I do this:<br>
><br>
> root@webserver01:~# modprobe ip_vs_dh<br>
> FATAL: Could not load /lib/modules/2.6.24-2-pve/modules.dep: No such<br>
> file or directory<br>
><br>
> Do I have to load the modules directly from the host?<br>
> or is there a nice how-to, tutorial to setup a loadbalacing apache<br>
> cluster using containers that you guys know?<br>
><br>
<br>
</div>i was about to ask if this was on the host or guest but you answered it<br>
for me :) module loading needs to be done on the host; openvz is a<br>
shared-kernel system (meaning the guests share the host kernel) so from<br>
a security standpoint, it doesn't make much sense to allow the guests to<br>
have control over the host's kernel/modules.<br>
</blockquote><div><br>Actually I've loaded the modules in the host and the result is the same:<br>proxmox:~# modprobe ip_vs_ftp<br>proxmox:~#<br>proxmox:~# lsmod<br>Module Size Used by<br>ip_vs_ftp 15492 0<br>
ip_vs 105696 2 ip_vs_ftp<br>kvm_intel 57960 0<br>kvm 191752 1 kvm_intel<br>vzethdev 23808 0<br>vznetdev 32776 4<br>simfs 14320 2<br>
vzrst 155688 0<br>vzcpt 129976 0<br>tun 23168 2 vzrst,vzcpt<br>vzdquota 58864 2 [permanent]<br>vzmon 58520 6 vzethdev,vznetdev,vzrst,vzcpt<br>
vzdev 13064 4 vzethdev,vznetdev,vzdquota,vzmon<br>xt_tcpudp 12288 0<br>xt_length 10752 0<br>ipt_ttl 10752 0<br>xt_tcpmss 11264 0<br>xt_TCPMSS 13568 0<br>
iptable_mangle 13824 2<br>iptable_filter 13696 2<br>xt_multiport 12288 0<br>xt_limit 12032 0<br>ipt_tos 10496 0<br>ipt_REJECT 13952 0<br>ip_tables 33256 2 iptable_mangle,iptable_filter<br>
x_tables 33672 10 xt_tcpudp,xt_length,ipt_ttl,xt_tcpmss,xt_TCPMSS ,xt_multiport,xt_limit,ipt_tos,ipt_REJECT,ip_tables<br>ipv6 342016 41 vzrst,vzcpt,vzmon<br>
bridge 73128 0<br>dm_snapshot 28256 0<br>dm_mirror 34432 0<br>e1000 176068 0<br>thermal 27168 0<br>button 18336 0<br>intel_agp 38304 0<br>
processor 49768 1 thermal<br>evdev 22912 0<br>pcspkr 12288 0<br>scsi_wait_scan 10112 0<br>dm_mod 79736 9 dm_snapshot,dm_mirror<br>usbhid 43616 0<br>
hid 52544 1 usbhid<br>usb_storage 90304 0<br>libusual 31072 1 usb_storage<br>sd_mod 40448 3<br>sr_mod 27684 0<br>ide_disk 26496 0<br>
ide_generic 9856 0 [permanent]<br>ide_cd 43040 0<br>cdrom 48936 2 sr_mod,ide_cd<br>ide_core 144152 3 ide_disk,ide_generic,ide_cd<br>shpchp 45596 0<br>
pci_hotplug 43312 1 shpchp<br>uhci_hcd 37408 0<br>ehci_hcd 48908 0<br>usbcore 178608 6 usbhid,usb_storage,libusual,uhci_hcd,ehci_hcd<br>iTCO_wdt 22992 0<br>
iTCO_vendor_support 13188 1 iTCO_wdt<br>i2c_i801 19740 0<br>i2c_core 36352 1 i2c_i801<br>ata_piix 31492 2<br>pata_marvell 16768 0<br>pata_acpi 17152 0<br>
ata_generic 17412 0<br>libata 184496 4 ata_piix,pata_marvell,pata_acpi,ata_generic<br>scsi_mod 187192 5 scsi_wait_scan,usb_storage,sd_mod,sr_mod,libata<br>isofs 47144 0<br>
msdos 19712 0<br>fat 67760 1 msdos<br><br>root@webserver01:~# modprobe ip_vs_ftp<br>FATAL: Could not load /lib/modules/2.6.24-2-pve/modules.dep: No such file or directory<br><br>any other hint?<br>
<br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
if you would like to avoid this, Xen allows for private kernels i<br>
believe (as long as you're using paravirtualization.. can a Xen guru<br>
correct me if i'm wrong on this?)<br>
<div class="im"><br>
<br>
> googling around found this:<br>
> <a href="http://lbvm.sourceforge.net/" target="_blank">http://lbvm.sourceforge.net/</a><br>
> which only works under red hat clustering services<br>
> is there an alternative for debian/ubuntu?<br>
><br>
> my host is a proxmox(debian) machine using several ubuntu/debian<br>
> containers and VMs<br>
><br>
> TIA<br>
><br>
<br>
</div>looks like it is supported pretty well- per the homepage, "The LBVM<br>
consists of several scripts that allow to load balance virtual machines<br>
(**currently preconfigured: Xen and OpenVZ**) among physical servers -<br>
the algorithm is fully configurable. LBVM uses the Red Hat Cluster Suite<br>
to provide high availability and rgmanager (part of the Red Hat Cluster<br>
Suite) to perform the actual migration." (emphasis added)<br>
</blockquote><div><br>It is a nice tool for a rhel/centos but I will still love to use in debian/ubuntu though.<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
-----BEGIN PGP SIGNATURE-----<br>
Version: GnuPG v2.0.10 (GNU/Linux)<br>
Comment: Using GnuPG with Mozilla - <a href="http://enigmail.mozdev.org" target="_blank">http://enigmail.mozdev.org</a><br>
<br>
iEYEARECAAYFAkm/8/EACgkQ8u2Zh4MtlQradgCgrRBnm8SWnlEdP9IP9ozjpP2/<br>
qv4AoKIFrACXIqnXV08cf06tmBR1yhYK<br>
=rwWI<br>
-----END PGP SIGNATURE-----<br>
</blockquote></div><br><br clear="all"><br>-- <br>"It is human nature to think wisely and act in an absurd fashion."<br><br>"Todo el desorden del mundo proviene de las profesiones mal o mediocremente servidas"<br>