On Saturday, I noticed that snapshot 20160205 had been released. So I downloaded that for this month’s install.
The install itself went quite smoothly. I installed the KDE (i.e. Plasma 5) desktop. Additionally, I installed networkmanager-gnome, so that I could test that with KDE.
My install was into the same external drive that I have previously used. It was connected to my laptop. My install was into an encrypted LVM. I imported the mount points during the partitioning section of the install.
I noticed two problems. The first was a repeat of the problem that I had n noted for my January install. Namely, the entry added to “/etc/crypttab” was bad. I had to go into rescue mode to repair that. But this was expected. That bug report is still open. So the problem was not a surprise and I knew how to deal with it. On this occasion, I used the UUID of the LUKS partition in “/etc/crypttab”.
My previous post was on setting up an NFS server. In this one, I will describe how I setup an NFS client.
My first decision was to use “autofs” (the automounter). With “autofs”, the NFS network share is not mounted until there is an attempt to access it.
Without autofs, you can get into a situation where you cannot boot the client computer unless the NFS server is running. So I prefer auto-mounting to avoid that kind of difficulty. Read More…
Last week I posted about sharing updates via NFS. In this post, I’ll describe how to setup an NFS server with opensuse. I’ll later add a post about setting up an NFS client. Both are relatively simple.
To setup the NFS server, I needed to install “nfs-kernel-server” and “yast2-nfs-server”. Both are on the install DVD and both are in the standard repos. When I first did this, I install “nfs-kernel-server” during my original install of the system. Then when I tried to configure, I noticed that “yast2-nfs-server” was missing, so I installed that.
Since getting IPv6 access, I’ve spent a little time looking into IPv6 addresses. So I thought I would share some of that on this blog.
An IPv6 address is 128 bits in length. So there won’t be a shortage anytime soon.
An address can be thought of as having three parts. There’s a prefix, typically assigned by the ISP. Then there’s a portion that can be used on a campus WAN, to distinguish different local subnets. And then there’s the local part which identifies individual computers on the LAN. The local part is often the last 64 bits of the IPv6 address.
In my case, my ISP has assigned a 60 bit prefix. My router has assigned the next 4 bits, which presumably would allow up to 16 subnets. And the last 64 bits are for the LAN portion, and are to be assigned to individual computers on the LAN. Read More…
Many of the readers of this may already have IPv6. My ISP, AT&T, has been a bit slow. Two years ago, they said we would have IPv6 by the end of June. They did not make that date.
I’m using their U-verse service, with a 2Wire 3800HGV-B gateway router. On May 20, there was a discussion thread at dslreports, about a new firmware release. Someone reported that this new firmware provided IPv6 support. Checking my router, I was still on the older firmware. However, about 1 week later, I noticed that my firmware had been updated.
I went to the site to check IPv6 support for U-verse, and it said the router was now capable. I clicked the link to have it configured. And, shortly thereafter, my router was reporting IPv6.
My home setup had included a secondary router. This was IPv6 capable, but I was not sure how well the double router would handle IPv6. So, the next morning, I reorganized so that the computers are all directl connected to the 3800HGV-B.
In an earlier post, I reviewed NetworkManager. In this post, I shall describe the options available for configuring a network.
The starting point with either KDE or Gnome, is to click on the NetworkManager applet in the system tray. That should show a list of available networks. Then click on the network to which you want to connect. Read More…
NetworkManager worked pretty well in opensuse 11.4, but then they broke it for 12.1. This was mainly a design change from the NetworkManager developers. The effect was that, in 12.1, the root password was required for setting up a system connection, and with a non-system connection the network name was often invisible.
As a result, my advice in 12.1 was to use only system connections. During beta testing, 12.2 seemed to have some of the same problems, and this was confirmed by complaints posted at the opensuse forums.
That all changed a few days ago, when a NetworkManager update came through. So I thought it was time review how well the updated version works. For completeness, Yast software manager shows that I am using NetworkManager 0.9.4.0-5.13.1. Read More…
Now that I have a newer desktop, with plenty of free disk space, I have setup file sharing for the home network. My opensuse desktop is the server, with both Windows and linux clients.
I am using a separate disk partition for sharing. It is mounted as “/shared”. I am encrypting that partition (LUKS encryption). Of course, that might sound silly; why encrypt what I am sharing on the home network? Mainly, this is because hard disks eventually fail. And if the disk is encrypted, somebody scavenging the dump for old disks won’t be able to recover useful data. I do expect to sometimes backup systems over the network, so there could be private data on the shared disk. Read More…
As part of my testing of opensuse 12.2 Beta1, I have been checking out how NetworkManager works for WiFi connections. And it still has the same problems we saw with 12.1, namely it required the root password to do almost anything.
I decided that it was time to get my hands dirty, and see whether I could make some changes that would result in the system behaving more sanely. This post describes what I have done. I would appreciate feedback (comments) from readers about this. Read More…