Some notes on Solaris 10 U4 x86 as an iSCSI target
The latest release of Solaris 10 (S10U4 or 8/07 depending on how youlike to label it) has a built in iSCSI target implementation, and ZFSeven has integration with it so that it is easy to export space fromZFS pools. I've been poking it a bit, resulting in some things I wantto note down for my own later reference.
- Solaris 10 iSCSI target stuff is handled with the
iscsitadmcommand,which is helpfully only one letter away from the
Thinking through salts for passwords
No one who's sane stores actual plaintext passwords; instead theyhash the password in some way and store the result, throwing away thepassword itself as fast as possible. There are three ways you can dothis:
- you can hash just the password.
- you can hash some global value plus the password.
- you can hash some per-user value (a 'salt') plus the password.
If you hash just the password, anyone who can somehow gain access toyour password hashes can immediately run them through
A thought about the speed of IPv6 deployment
The following thought about the push for IPv6 deploymentstruck me recently, so I'm going to dump it here.
One of the problems with IPv6 is that its major benefit is to expandthe available address space, and the expansion of address space doesn'tbenefit people who are already on the Internet. Instead they are beingasked to do a great amount of work for the benefit of other people,which rarely goes down very well.
(The people who are already on the Internet by and large
Note to self: check for gigabit Ethernet
One thing I should remember is that before I start measuring things likeiSCSI performance, I should check to make sure that I'm actually using agigabit Ethernet connection. Conversely, if I get unexpectedly terriblehigh-level performance, perhaps I should make a point of immediatelychecking the network speed. Especially if the terrible performance isaround 10 megabytes a second.
(This is especially embarrassing when I forget this for the second time.)
This can happen to us for a combination of two reasons; first,
Dynamic rendering versus static rendering for websites
One of the long-standing splits for dynamic websites is between dynamiccontent rendering and static content rendering (where you use trickslike 404 handlers to generate a page'scontent only once, write it to a static file, and then afterwards justserve the static file).
(This assumes that static rendering is viable; for heavilypersonalized sites it may be more trouble than it's worth,if you are already effectively never generating the samepage twice anyways.)
Fundamentally the choice is tradeoff of
What may be causing my random NumLock issues
I think I may have finally woken up to a (potential) cause for my random NumLock issues : the xine media player. Oneindicator, one that I really should have paid attention to earlier, isthat often starting xine will turn on NumLock (and sometimes turn itoff immediately afterwards). It is of course hard to be entirely suresince the problem is erratic, but I've been doing some testing sinceI noticed the correlation and I don't think I've ever had the problemwhen
Jumbo frames on gigabit Ethernet on Solaris 10 x86
We've recently been looking into using jumbo frames on gigabit Etherneton Solaris 10U4 x86 (aka Solaris 10 8/07); it turns out that thisis more work than you might expect, and information on it is somewhatscattered.
First, as of S10U4 only a few gigabit network drivers are documented assupporting jumbo frames:
| driver | maximum MTU | chipset |
bge | 9000 | Broadcom, but only on some chipsets ;the manual says BCM570 |
The problem with big systems
In system design, we're periodically confronted with a choice betweenbuilding our environment out of a few big systems or a swarm of smallones. For example, consider storage; if you want 48 Tb of storage, youcould buy two Sun Thumpers or six or seven smaller 15-disk units. Bigsystems are often attractive; their very size gives them economies ofscale (and often means they cost less overall), and it's easier to dealwith only two boxes instead of six or
My problem with learning new programming languages
There's any number of programming languages I'd like to learn, rangingfrom the prosaic like C++, Java, and JavaScript (mostly because itwould make it easier to work with various interesting things writtenin them, like Firefox) to the esoteric favorites like Haskell andErlang. Unfortunately, I have a problem.
My problem is that I can't learning languages by themselves in theabstract; I can't just sit down and learn one just to know it. I needto be
The inconvenience of some DWiki design choices
When I designed DWiki back in the mists of time, I decided that itwould not require any writable data store (and certainly not an actualdatabase, because that would have required me to configure and run one);I wanted to be able to run it completely read-only, so I could worryless about its security implications. I also decided that it should nothave explicit metadata embedded in page data, which are all in regulartext files.
Blog entries have one primary time associated with them, their