[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [SAGE] Did everyone Migrate to LOPSA ?



On Wed, Nov 23, 2005 at 09:57:58AM -0500, Rodrick Brown wrote:
> I notice the amount of technical discussions has pretty dropped way
> down on this list, and its pretty much USENIX vs SAGE discussions
> going on, did all the technical folks move to LOPSA ?

 I personally haven't made up my mind yet.  I'm considering the
 following options:

    * Join LOPSA
    * Stick with SAGE
    * Do both and play it by ear regarding which one gets more of
      my focus
    * Ignore it all and just focus on other resources

 I'm going to base most of my decisions on what I hear and see at
 LISA next month.  That's where I hope to form a more complete
 understanding of what's going on than I've been able to glean from
 the discussions here.

 For the most part I suspect that many other "rank and file"
 lurkers on this discussion are doing the same; and I hardly consider
 myself to be of any particular status or influence.

 As for the lack of technical discussion here ... frankly I haven't
 been involved in any technical challenges for which this seems to be
 the appropriate venue.  Most of what I do these days is pretty tightly
 focused on obscure Linux internals and corner cases (and Red Hat
 deviations from mainstream sources in particular, especially as they
 relate to running older RHEL and legacy versions on newer x86_64
 hardware).

 I suppose I could offer up the following questions for those that
 which to spawn off some technical threads on them:

    * Has anyone here been trying to run RHEL3 on x86_64 hardware
      with 8 to 32 GiB of RAM? Do you also have stability problems?
      Anyone have a stable Linux x86_64 infrastructure on any 2.4 based
      kernels?  (RE: RHEL3 and 2.4 on x86_64)

    * Anyone using RHEL3 with a large number (50 to 100) autofs maps
      in LSF and/or Condor "grids?"  Any good documents on "best
      practices" for that arrangement?  (No fair saying that the
      "best practice" is to avoid NFS and autofs on that scale, esp.
      with 2.4 Linux kernels unless you can back it up with a pointer
      to credible, authoritative, detailed whitepapers or discussion
      on the topic. (RE: Linux and Lots of autofs maps)

    * How do people feel about NFS/TCP vs. UDP?  On Linux?
      (RE: Linux NFS/TCP vs UDP)

    * Anyone know how to get .tok files for biosconfig (IPMI tools)
      so that our imaging process can automate conformance testing
      and configuration of firmware settings?  The .tok token files
      are, naturally, BIOS specific and vendors seem to be treating
      them like they are irreplaceable trade secrets.
      (RE: IPMI biosconfig .tok files)

    * Anybody been using MediaWiki with LDAP authentication patches
      and perhaps some sort of page ACLs patches for internal
      collaboration and document management?  I'd love to use MediaWiki
      for this sort of thing; but having departments and teams working
      "in a fishbowl" in a 5000 person company is a pretty touch
      expectation for the particular corporate culture that I'm
      currently working in. (RE: MediaWiki with LDAP and ACLs)

    * Is there any sort of AJAX driven "file upload" facility that
      might allow users (even Windows users) to drop .doc, .ppt, and
      .xls files on a browser page and have it automatically uploaded,
      converted to HTML or wiki mark-up and posted somewhere?
      (RE: AJAX Drag and Drop)

    * How do people feel about Opsware for provisioning?  As a
      management agent?  Anyone have any experience with Opsware
      "shell?" (RE: Opsware)

    * What's the best resource on the 'net for X video drivers for
      Linux on laptops to drive CAD (and especially EDA CAD tools)
      on both the internal LCD panel and an external monitor or
      projector?  What's the best laptop, video chipset, and Linux
      distro for this application. (RE: Linux and Laptop Video Drivers)

    * What are people doing for laptops in their LDAP environments?
      Are most pople just reserving run-level 4 for on or off the 
      LAN/VPN?  Have people been using things like pam_ccreds and
      nss_updatedb? (RE: Linux and Laptop/Mobile LDAP)

    * In the Linux kernel on 32-bit systems; how are the PSE and PAT
      (Page Size Extension and Page Attribute Table?) used?  When and
      how is the system using 4MB pages rather than the normal 4KB
      page size?  Which versions of the kernel introduce these features?
      How do these relate to the hugetlbfs? (RE: Linux and 4MB Pages)

    * Are there any really good online or commercial training resources 
      for those who are undertaking a transition from RHEL to SLES
      (SuSE/Novell)?  (For IT on the one hand, and perhaps seperately
      for programmers and technical users on the other)
      (RE: RHEL to SLES Migration Training and White Papers)

    * Other than cfengine and the code from Steve Traugott's 
      infrastructures.org site; does anyone know of a framework or
      suite of simple shell scripts (no binary agents or daemons) that
      can test a variety of IT configuration/conformance details and
      optionally mitigate them?  (Things like: Does this host have
      a valid system hostname that's consistent with it's DNS and
      reverse DNS settings?  Does it have a proper and working NTP
      configuration and is its time within a reasonable skew of some
      given time reference?  Is sshd running and properly locked down?
      Do we have an sshd "watchdog" (supervisor?) or is it running
      from inittab under a respawn directive?
      (RE: Lightweight IT Conformance Testing and Mitigation)

 Those are just a few of the technical issues that I'm currently working
 with in some way or another.  (If anyone reading this does choose to
 spawn off threads on any of these topics please consider changing the
 subject and perhaps even using the ones that I've suggested above.
 Also please consider spawning off separate threads for each).


-- 
Jim Dennis