Every so often, someone writes a program that is overly chatty bydefault (usually in the interests of being user-friendly) and thenthinks 'I know, I'll make sysadmins happy by giving it an option to bequiet'. They are then often surprised when sysadmins seem to find thenew option not all that useful or compelling. The particular exampleI'm thinking of here is Debian's apt-get , but there are others.

I can't speak for other sysadmins, but

I've seen a reaction or two to my entryon ZFS deduplication that suggestthat the reason ZFS dedup has significant issues is that the Solarisprogrammers did a bad job on it. I want to set the record straight: thisis false. Deduplication is fundamentally not an easy operation and noform of deduplication can be fast unless it is fed huge amounts ofresources.

Any deduplication system needs to maintain a dedup table in someform. It's the core data structure because the core operation indeduplication is '

I recently read Die, VPN! We're all "telecommuters" now (via Hacker News), which prompted me to think about why we have a VPN server for external access andwhy we're likely to keep it for the foreseeable future. What it boilsdown to is two factors.

The major reason why we have a VPN because of limitations. We have aVPN because of insecure internal software that we can't expose to theInternet, because of a presumed lack of security of some of

A commentator on my previous entry on enticing ZFS features suggested that ZFS deduplication (coming inSolaris 11) would be one of them. Unfortunately it's not; ZFSdedup is not something that we or any number of other sites aregoing to be able to use in the near future, and probably notever. The core problem is its impact on performance, management,and system requirements for ZFS servers (all of them are bad).

Because this is somewhat long, I'm going to put

I've written before about how our fileserversare basically appliances and so don't patched because we don't liketaking any risks of destabilizing vital core services. Today they runwhat is more or less Solaris 10 Update 8 (which causes some problems),and today I've gotten interested in inventorying what features fromsubsequent Solaris versions might be attractive enough to us to cause usto upgrade.

(Note that there are a lot of ZFS features in S10U9 and S10

There are a number of things about DWiki that Iwould change if I was writing it from scratch again. I'm not entirelysure that I'd completely replace its dialect of wikitext with a standardone such as Markdown, but there are certainly aspects of it that I wouldredo. In particular, one part of DWiki 's wikitext dialect has turned outto be a terrible mistake.

You see, when I was designing my wikitext dialect ,for some reason I decided that digits at the start of line

I'm in the process of upgrading my office workstation from Fedora 14to Fedora 15 (since I've already piloted Fedora 15 on my home machine).I'm doing it with a yum upgrade , sincethat's the sane way to do it if you don't want to be without yourmachine for hours. And I got the following (fatal) error:

warning: rpmts_HdrFromFdno: Header V3 RSA/SHA256 Signature, key ID 06

We have some bookable compute servers, machines that can be reservedfor a single person. When we first introduced them and for a long timeafterwards, demand for them was low and we handled all of the bookingand related tracking by hand (mostly with email). Demand has beenpicking up lately and as a result, we've automated much of the process.

(The trigger for automation was when the machines actually got queuesof people waiting to reserve them. Handling a single active booking permachine was relatively

When you're storing encrypted passwords, you can do two different thingsto ruin the life of an attacker who got your password hashes, or atleast annoy them; you can salt your passwords in various ways and youcan use a slow password hashing function. These two have different butcomplementary effects.

(I've talked about the effects of the various ways of salting passwordsback in SaltingPasswords .)

Salting passwords reduces the payoff an attacker gets for makinga single password guess. They slow down how fast an