[nycbug-talk] Advanced UNIX Basics Management

David Lawson dave at donnerjack.com
Wed Sep 23 13:21:27 EDT 2009


It's worth looking at what version of systat you've got installed and  
definitely worth thinking about upgrading to the latest (or later)  
ones, at least on linux 2.6 kernels you can get fun stuff like per  
process disk IO statistics and all kind of niftiness out of it.

--Dave
On Sep 23, 2009, at 12:47 PM, Isaac Levy wrote:

> Hi All,
>
> Much like the 'Tips-And-Tricks' thread a few weeks back:
> At work, I'm charged with giving a talk to Developers and DBA's on the
> following topic:
>
>   "UNIX Process, Memory, and Disk- Userland Monitoring Tools for Non-
> UNIX developers"
>
> My company is a Java shop, with plenty of Ruby, Python, and loads of
> shell scripts running around- in BSD and Linux systems- so this talk
> is all about trying to get everyone on the same page with 'the classic
> basics'.
>
> --
> With that, I thought I'd hit list to see what folks would have any
> input?  What am I missing?
>
> The high-level outline is below, any comments/criticism is welcome,
> I'm looking for stuff I've missed-
>
> Best,
> .ike
>
>
>
> --
> Each section below is split into a 'read whatever is happening' part,
> and 'do something with whatever is happening' part- most devs' tasks
> at my company just need to have visibility into things like what part
> of their code is eating the system, basic issues.
>
> I am explicitly *not* looking for good 3rd party tools, (pstree, for
> example)- I am looking to cover the basics of what's just expected to
> be there on our typical stock UNIX systems- (FreeBSD and OpenBSD, and
> CentOS Linux here, to be precise).
>
> I'm also not really looking for DTrace type tools, that's a whole
> exploration on it's own- especially when it comes to apps which aren't
> written in C.
>
>
> ##############################
> - Userland/Kernel Structure Basics (2 minute spiel)
> - man(1) is your friend, so is dmesg(8)
>
> - Processes
>   - stats/info facilities
>     + using procfs(5)
>     + ps(1) (flags and some handy awk(1) parsing)
>     + top(1) (briefly, everyone knows top...)
>   - management tools
>     + kill(1), killall(1) (flags!)
>     + nice(1), renice(8)
>
> - Memory
>   - stats/info facilities
>     + ps(1) (flags and some handy awk parsing)
>     + top(1) (briefly, everyone knows top...)
>     + swapinfo(8)
>   - management tools
>     - swapon(8), swapoff(8)
>
>
> - Disk
>   - stats/info facilities
>     + iostat(8)
>     + df(1) and du(1)
>     + lsof(8) (non-stock on many UNIX systems, but worth mention?)
>     + top(1) disk i/o tricks
>   - management tools
>     + disk mount(8) basics
>     + nfs, living with it basics
>   - advanced but very useful for developers:
>     + memory filesystems (creating, using)
>       - disk-backed memory filesystems
>
> --
> Bonus Networking section, perhaps,
>
> - Network
>   - UNIX stats/info facilities
>     + ifconfig(8)
>     + netstat(1)
>     + tcpdump(1)
>   - UNIX management tools
>     + ifconfig(8)
>     + netstat(1)
>
> ##############################
>
>
>
> _______________________________________________
> talk mailing list
> talk at lists.nycbug.org
> http://lists.nycbug.org/mailman/listinfo/talk
>




More information about the talk mailing list