May 2006 Archives

On laptops

AnandTech recently did a review of a laptop I recently purchased. The Dell Inspiron E1705. I had a 25% off coupon, and was looking for a new laptop anyway. You can read the review here.

One of the quibbles they had was that the keyboard was too small. I have to agree. One of the prime reasons I got the 17" version instead of the 15" version was the mistaken belief that the keyboard would be larger. Oops. This keyboard is the same size as the 15" one, and that will give me problems. My old laptop had a smaller keyboard, and it was on that laptop that I learned I can't type for hours at a time on something that small without side-effects. I can't write a novel on this laptop's keyboard. At least, not in anything resembling 'reasonable' time.

The other thing that tripped me up was the screen-size. Like most laptops these days, this is a wide-screen unit. And a 17" wide-screen somehow looks smaller than a 17" standard. By golly it is shorter, which I've noticed! For what I do with it, I'd have preferred a standard 17". But if Dell could have put in a larger keyboard in that widescreen body, I'd live with the wide-screen ;).

Because I won't be gaming with this rig, I got the embedded graphics solution. This has a considerable effect on battery-life as it turns out. The tests are shown here. I've found that the battery I got with this laptop actually is good for 6 hours. 6 hours! My old laptop was good for about 3.25 hours on a brand new battery. That uptime is just for 'office applications' not DVD playback (old laptop was too elderly to do this...dvd...thing. It didn't even have a CD-burner).

The only personal problem I've had with the unit is that the backspace key falls off. There is a simple fix for this, as the Dell tech said. Perhaps my early keyboarding on an IBM-XT 'clicky' keyboard makes me mash keys w-a-y harder than they need to. Or perhaps it is a design fault. Who knows. But that key keeps falling off, though not in the past week.

Other than the one fault, the unit is solid. I hope to repartition and get a SUSE partition at some point, but I'm not quite there yet. First I need to figure out how to image it. That'll take some time.

Changes a-coming

Since our fiscal year slash budgetary period starts in July, we get a lot of summer projects as a result of new funding. Several things were passed out today.
  1. We will be changing our Anti-Virus /Anti-Spam solution away from McAfee. I can't say who it is, but this is a migration and therefore a Project. The solution that won did so by having a much better anti-spam product. I was not involved in the selection process.
  2. The Distaster Recovery SAN was approved. This'll go into the Bond Hall datacenter. It'll be an HP MSA, rather than an EVA, so we'll have to do some creative engineering to set up the data mirrors. This'll be fun. That space will also be used for backup-to-disk for some undefined instances. Policy to be set later once we have the tech on-site and have played with it. This DR SAN will be mirroring the NetWare cluster, the Exchange cluster, and the yet-to-be-built BlackBoard cluster.
  3. We'll be getting a snazzy new tape library to support the added space in that DR SAN. Mmm. Backup hardware.
  4. The project to come up with a fully enterprise Backup Software Solution (we don't have an internal standard) is... continuing negotiations. NetBackup is the software of choice, but they're being unreasonable when it comes to pricing.
  5. A Student Tech Fee proposal was approved, so we'll be carving out a 2TB hunk of the SAN for GIS data. There will be a server attached to this to handle the software end of it.
  6. We'll be getting a quote from HP about upgrading our EVA to something that can handle:
    1. Drives larger than 30GB.
    2. More than 4 drive shelves.
Fun times this summer. And I'll be gone a goodly percentage of July, and a week-and-a-bit in August. Summer upgrades are great! You get to do them around vacations and things, so you're already shouldering slightly higher normal workloads, and have to do all these other projects too. Plenty of high-value thingies on that list. And really, playing around with $100,000+ hardware really makes this job worth it some days.

Tags: , , ,

Nifty Microsoft tool

I found a link to the Microsoft Standard User Analyzer. This tool was released on the 23rd. According to the site, it is designed to check applications out for Administrator dependancies. This will be quite useful for desktop-image builders and other such people who get to figure out how many apps are forcing users to run as Admin this week. Vista reportedly has better priv-separation features in it, but XP and such don't.

Interesting tool. I won't get a lot of use out of it, but I know folk who will.


An exciting week

Yesterday at about 9:30, I noticed that the master eDir server was throwing memory allocator errors. I've seen a bajillion of those since SP3 a year and a bit ago, so that's nothing new. So was the server refusing to shut down gracefully. This has been a common problem with our blade servers, and I haven't figured out where the hang is. It stuck there for a good long while and CPU1 never was shut off, so I hard-booted the server. No other way it was coming back, and in its state it was 'sort of up' and causing problems in computer labs and some logins. So it HAD to reboot.

Unfortunately, it didn't come up right. eDir was in some strange state when the reset button was pressed, and it was unrecoverably corrupted. Never seen that before. So we had to reinstall eDir on that machine from the 'down server' TID. Owie.

It also held our CA, and the backup I thought I had didn't exist.

Double Owie.

It also held the masters for most of the replicas, and was in all the replica rings.

Triple Owie.

It was one of our two SLP DA's, and clients out there had a hard time with the data it was giving, when it was giving it.

Quadruple Owie.

It was a long day yesterday.

In one of those 'silver lining' things, the outage showed to us how many LDAPy things rely on that specific server for service. Now that we have LDAP load-balanced by an F5 BigIP, this was the main chance to get folk pointing at the server directly to point at the BigIP's virtual IP. Yay! That way, that IP will stay up so long as at least one of the three eDir servers is up. No repeat of yesterday.

Tags: ,

Changing bosses

My grand-boss is retiring in July. Since he is effectively our CIO, this has a wide-ranging effect on our future. The search process has come down to three candidates, and all three are showing up over the next 7 days for meetings with staff.

This is somewhat unusual to my experience. Managers, especially in civil service like we are, get hired and the top candidate is selected. All without input outside of the search comittiee. Getting a voice in the process of selecting my grand-boss is a little... weird.

Being civil service, certain things can be taken for granted. A series of interviews were held, duringwhich the candidates were assessed against a pre-defined set of parameters. After the interview process was completed, each candidate was rated against the parameters and a list was drawn up, first to last, of how they placed against the list of parameters. The top three candidates were then offered to the hiring manager (a Provost).

At least, that's how it is supposed to go. I have no doubts at all that the selection comittiee has a top candidate in mind already. Further more, I have no doubts that the staff interviews are not likely to have a significant impact on the person in the top slot. There is always the chance that some significant and objectively useful detail will come out in the interviews with future-staffers and that'll affect the rankings, but that is not a given. I also doubt that a report of, say, "all of ATUS hated the guy," would have little impact as well, given the objective nature of civil service hiring rules.

The prospect of a new CIO is giving the existing CIO a case of short-timer syndrome. Involuntarily. There have been a few major decisions made of late that the various people directly under the CIO are quietly stonewalling implimentation, prefering to await the new guy's (or gal! One of the candidates is actually female) opinion on the whatever. I suspect this is a very common thing when a high level exec leaves in a planned way such as this. The outgoing exec is going to be sticking around in some undefined capacity, so the chances of the new person changing stated direction are slimmer than some would hope. IMHO, of course.


NDPS security fault

I saw some of the patches for this a few days ago, but the full details of the fault have been posted by Hustle Labs. There is a integer overflow that is exploitable. What's more, the overflow is predictable enough that exploit code can be written to make it more than just a denial-of-service. Not a real issue for NetWare servers (I haven't seen a root-kit for NetWare yet), but it will be an issue for Novell Client for Windows users.

Tags: ,

Trending things

Ran into a feature of google I hadn't heard of before, trends. It tracks search-term trends. So I threw in a few things just because.

Pretty steady, not too surprising. What was somewhat surprising is that Germany is the #2 region hitting that. Interesting.

Owie. Steady downslide. The top search language was Russian, which is interesting.

Also a steady downslide, but not as steap as NetWare.

Pretty steady, actually, with some down-slip lately. Sharp uptick very recently, which I expect is due to SuSE 10.1 releasing.

Steady increase in volume. Good to see!


The good of blogs

An anonymous poster suggested the use of a program I completely missed called 'fileop' that exists in the Iozone distribution. From the source:
 * Usage:  fileop X
* X is a force factor. The total number of files will
* be X * X * X ( X ^ 3 )
* The structure of the file tree is:
* X number of Level 1 directorys, with X number of
* level 2 directories, with X number of files in each
* of the level 2 directories.
* Example: fileop 2
* dir_1 dir_2
* / \ / * sdir_1 sdir_2 sdir_1 sdir_2
* / \ / \ / \ / * file_1 file_2 file_1 file_2 file_1 file_2 file_1 file_2
* Each file will be created, and then 1 byte is written to the file.
Which happens to be a great way to test whonking huge directory trees. Balanced trees, of course, but still whonking huge. We see a problem with directories involving thousands of directory-entries, and this tool could be used to see if their behavior changes between when viewed over the network and when viewed locally.

Unfortunately, my test blade is currently being used as part of the BlackBoard upgrade project so I won't get that back until June. But I plan on doing some tests to see how nss vs. reiser vs. reiser w/separate journal behaves.

More zen stuff

The most recent novell-client push we did included a Zen Agent push. This is the first step in a Zen7 upgrade, and that end of it went great. No worries.

Except. Except the transform file for the Zen Agent included a pair of registry keys that forced all workstations that installed the client to register as a specific workstation. Sub-optimal. I've fixed the transform file, but I suspect some of our desktop support people copied the install directory to CD for use in offline installs, and included the bad MST. Grr.

I've got like 95% of the workstations registering under their own names now. But that last 5% is proving insidious. My most recent Zen-Fu attempt is to associate an application object that runs "zwsreg -unreg" as the System user. This forces the workstation to cleanly unregister. Then, as a side-effect, shortly after that (or on the next reboot or whatnot) it'll query the Name Resolver to figure out if there is a "zenwsimport" that resolves to an IP address. Which will work, since I have it in WINS right now. So things are re-importing. Since doing this, I've gotten something like 150 new workstation registrations and our total is around 3575 machines.

Next week I'll push out another force-run application that'll do "zwsreg -importserver [ipofimportserver]". Which should catch the remainders.

Tags: ,

Testing Wiki

| 1 Comment
I downloaded the snipsnap wiki software. I was looking for a Wiki software I could run on my workstation to experiment with data organization. I have another project at home that I am trying as well. This is a wiki driven entirely in Java, and so far it is working pretty good.

Though, I do wish I could get something as featured as, say MediaWiki, but that is a 'thick' Wiki client. It could be installed to my local machine, but it would require a php/tomcat/mysql/apache stack to make work, and that's just too much software for what I'm going to use. This package seems to be a nice compromose.

We shall see how well it works out.

Tags: ,

SLP problems

I had noticed lately that rconip wasn't showing as many servers as it should. After I did the round of libc and openssh updated Tuesday, all of the WUF servers were absent from the list. Poking around showed that the WUF servers were not discovering the DA's. I wasn't sure how this could be, since SYS:ETC/SLP.CFG was correctly setup, and we had the right scope.

Turns out that my SLP DA Discovery Options parameter was not set correctly. It was set to DHCP-only, which that particular subnet has no DHCP. Setting it to query the static file and use Multicast (which we have turned on) got things reporting again.

I wonder how it got unset?

Tags: ,


We have swinging items here in the office that aren't related to wind. No quakes in the area, but there was an 8.1 in the pacific basin an hour ago. Could this be related? The Mt. Baker webicorder seems to show a spike at about 15:39 UCT, which the theoretical P-Wave chart for the quake says is within a minute of expected. What's more, the webicorder chart says that it is still seeing shaking as of this minute, and my blinds are still swinging. Am I seeing the side-effects of a quake half way around the world??


Wee hours update

Last night I spent time getting a LibC and OpenSSH update into the WUF machines. This should, in theory, make things more stable. Certainly on the SSH front, since there are sigificant tweaks to that package to make it work better on NetWare. And LibC updates only make me happy.

Unfortunately, one of the StuSrvs threw 111 errors when it was the MyWeb server. Don't know where THAT came from, time for checking when I get in again.