I'll start with a story.

I tweeted some screenshots from our Prometheus setup recently.As it happens, there's something interesting going on in the firstone, which you can find a 1920x1010 'full sized' PNG of here . Thisis a dashboard of one of our Amanda backup servers during backupsand thus as it writes data to the disk drives we use as 'tapes' . If you look at the 'disk write bandwidth' graph,you can

In the Hacker News discussion on my entry onfinding a bug in GNU Tar , theregrew a sub-thread about the many different cases of read() 'sreturn value . Thisis a good example of the potential complexity of the Unix API inpractice, and to illustrate it I'm going to run down as many of thecases that I can remember. In all cases, we'll start with ' n =read(fd, buf, bufsize) '.

The simplest and most boring

These days I have a HiDPI display at home ,although work is still dual Dell U2412Ms (1920x1200 at 94 DPI).This has made me run into an unusual problem, which is sort of theironic inverse of the HiDPI tiny app issue .Put simply, every so often I want to take a screenshot of something,whether it is how xterm looks or some web pages in Firefox . If I makethese screenshots at home, the natural

For reasons beyond the scope of this entry, I've recently beenlooking at ARC stats, in part using a handy analysis program byRichard Elling .This has gotten me looking again at ZFS (k)stats related theprefetching, which I touched on before in my entry on some basicZFS ARC statistics and prefetching . So hereare some notes on what I think these mean or might mean.

To be able to successfully prefetch at all, ZFS needs to recognizeand predict your access pattern. The extent to

One of the things that I've been noticing these days is how muchplaintext parts of emails are fading away. I'm not talking hereabout HTML-only emails (which have been on the rise here for years);instead, this is about MIME multipart/alternative email whichtheoretically has both a plaintext and a HTML portion. For yearsI've had my mail system set to show methe plaintext version instead of the HTML version. For a long timethat worked reasonably well, but increasingly

Every so often something happens in my work that makes me think,even if I don't know what conclusions to really draw from it. Irecently mentioned that we'd found a bug in GNU Tar, and the storyof how that happened is one of those times.

We back up our fileservers through Amanda and GNU Tar. For a long time, we've had a problemwhere every so often, fortunately quite rarely, tar would freak outwhile backing up the filesystem that held /var/

A lot of the time, moving to a new Fedora version is a smooth processfor me, where everything basically works the same as the previousversion but the software versions are more recent and up to date.Sometimes there are new tweaks that I have to make or changes thatI have to adapt to; for example, when I went from Fedora 25 toFedora 26 I wound up with font rendering issues and eventually decided to switch to standard Fedora font renderingand fonts . Based on my lack of notes

Suppose, not hypothetically , that youhave just performed an operating system upgrade that upgraded theversion of xterm . On the surface, this version of xterm looksjust like the old one; for example, your default xterm looks likethis, which is normal and how it used to.

However, you also have 'root xterms' for your su sessions, witha red foreground to make them stand out as special ( cf ). When you call one of these up forthe first time, it looks like this

The big news today is contained in an email thread called Thefuture of ZFS in FreeBSD (via any number of places), where the crucial summary is:

[...] This state of affairs has led to a general agreement among thestakeholders that I have spoken to that it makes sense to rebaseFreeBSD's ZFS on [ZFS on Linux]. [...]

This is not a development that I expected, to put it one way. Allan Jude had some commentary on OpenZFS in light