1. 26 Apr, 2004 4 commits
    • Robert Ricci's avatar
      Some big changes to assign, and some related changes to assign_wrapper · 238dce73
      Robert Ricci authored
      and ptopgen.
      
      Add link typing to assign. Each virtual link is given a single type.
      Each physical link is given one or more types. A virtual link will
      only be mapped to a physical link which can satisfy its type. In both
      the top and ptop files, the link types are now mandatory, and they
      fall at the end of the mandatory link arguments.
      
      This differers from the 'regular' type system in two ways. First, a
      plink is not constrained to filling only one type at a time. If we are
      using emulated links, a plink could satisfy, say, an 'ethernet' link
      and an 'fxp' link at the same time. This seems to more naturally match
      the way we'll use link types.  Second, there are no counts assoicated
      with link types, as there are for node types. ie. a link is not an
      'ethernet:1' link, it's an 'ethernet' link. Presumably, when
      multiplexing virtual links onto a physical one, it's bandwidth that's
      the factor that limits the multiplexing.
      
      The link type is now taken into account when constructing pclasses,
      and in the mapping precheck.
      
      As a side-effect of these changes, the silly 'count' argument on the
      end of link lines in the ptop file, which was used for the fake LAN
      nodes, is no longer supported.
      
      The implementation could be a bit more efficient, but that would mean
      tossing more of the stuff we do with boost's graph library. I think
      this should happen, but today is not the day.
      
      Modify assign_wrapper and ptopgen to spit out top and ptop files in
      the new format.
      
      Changed the constant LINK_UKNOWN to LINK_UNMAPPED - the new name more
      accurately reflects the way this constant is used.
      
      Add a new '-n' flag that tells assign not to do annealing, just exit
      after the type precheck.
      
      Clarify the usage message for the -c flag.
      
      Removed some dead code for dealing with LAN nodes.
      238dce73
    • Mike Hibler's avatar
      Cleanup Makefiles: · 297019fb
      Mike Hibler authored
      1. "make clean" will just remove stuff built in the process of a regular build
      2. "make distclean" will also clean out configure generated files.
      
      This is how it was always supposed to be, there was just some bitrot.
      297019fb
    • Mike Hibler's avatar
      Added config option DISABLE_EXPORTS_SETUP for sites without proper control · 56acaee5
      Mike Hibler authored
      of their file server (aero)
      
      Extend DISABLE_NAMED_SETUP: when set, we don't set a nodes hostname to
      <name>.<eid>.<pid>.<domain> since that won't resolve.  Just stick with
      pc<XXX>.<domain> in those cases.  The various sethostname* scripts are
      now .in so that they get preprocessed to check for the option.
      56acaee5
    • Leigh B. Stoller's avatar
      Changes to exit status stuff to reflect recent changes by Rob to how · 1c4a613c
      Leigh B. Stoller authored
      assign exits (exit codes).
      
      * in assign_wrapper, no longer return any status from assign to the
        caller. This was pointless. Instead, return 0 on success, 1 on
        controlled error, and -1 on uncontrolled error (die() called
        someplace). Add in CANRECOVER bit whenever the wrapper exits, even
        if uncontrolled, by putting in an END block to catch the die. This
        should prevent certain cases where a swapmod error would be flagged
        as not recoverable.
      
      * Remove most of the assign output processing since we no longer
        return its codes. Still print a portion of it to the log though.
      
      * Change call to fatal() in assign_wrapper; do not pass an exitcode
        since in every case it was the same damn thing!
      
      * Change tbswap to no longer carry assign_wrapper exit code to its
        exit.
      
      * Change the batch daemon to treat all errors as continuable (keep
        batch queued) unless exit code is -1. We will need to revisit this a
        bit perhaps, when Rob adds precheck code.
      1c4a613c
  2. 23 Apr, 2004 5 commits
    • Mike Hibler's avatar
      Back out seting of vfs.nfs.bufpackets. · e1e6d268
      Mike Hibler authored
      It didn't fix the problem and it makes nodes fail on startup sometimes.
      e1e6d268
    • Kirk Webb's avatar
      Fixed minor typo in __addNode() · a8129f1f
      Kirk Webb authored
      a8129f1f
    • Leigh B. Stoller's avatar
      I have changed xmlconvert to do the following (note that I have not changed · a3ad59cd
      Leigh B. Stoller authored
      the parser; it still spits out the simplistic xml stuff and xmlconvert
      still reads that):
      
      * When generating an XML representation of an experiment, I read the
        virtual tables and build a *flat* data structure.
      
      	{"settings" => {pid:testbed, eid:foobar ....}.
      	 "nodes"    => [ {vname:nodeA, osid:RHL-STD ... },  ],
      	 "lans"     => [ {vname:link0, member:nodeA ... },  ],
               ...
      
        So, nodes is a list of node hash tables, lans is a list of hash tables,
        etc. All of the other virt_tables are represented as well.  You can find
        everything you want as long as you the names of the fields!
      
      * Once I have the above data structure, I use the XML-RPC lib to turn it
        into an XML document, and spit out the resulting XML string (a very long
        string!) Note that XML-RPC has its own format that allows you to convert
        data structures into XML and back; thats how you issue an RPC from your
        python client to a perl server.
      
      * Our python XMLRPC server basically uses xmlconvert as a backend so it can
        ship the above data structure back to client. We could also allow the
        client to ship the data structure back, and have xmlconvert upload the
        changes back to the DB.
      
        The reason for doing it in xmlconvert as a backend is so that this XML
        stuff is in just one place; whoever reads/writes this stuff has to know a
        lot of internal detail about an experiment (like what things should not
        be exported and what things we should not allow to be imported). I prefer
        to keep that all in one place since its easy to pass XML strings around
        (as long as everyone is talking the same dialect).
      
      Future work is to pick an XML representation that is not tied to
      XML-RPC. Ideally, we want to take a data structure in Perl, spit out
      the XML tree (with a DTD) and be able to reconstitute that in another
      program, using the XML Parser library (based on expat, which is common
      to Perl, Python, C).
      a3ad59cd
    • Mike Hibler's avatar
      Change watchdog policy so that all plab nodes report their host keys. · 1befb964
      Mike Hibler authored
      Reduce amount of logging for malformed host keys, we are going to be getting
      a lot of those til the new watchdog makes it out to all remote nodes.
      1befb964
    • Mike Hibler's avatar
      Add distinct options to use NFS for loopback mounting of local or remote · a6b2f510
      Mike Hibler authored
      filesystems.  Also add sysctl call to increase socketbuf size for NFS
      sockets.
      a6b2f510
  3. 22 Apr, 2004 8 commits
  4. 21 Apr, 2004 11 commits
  5. 20 Apr, 2004 5 commits
  6. 19 Apr, 2004 4 commits
  7. 18 Apr, 2004 2 commits
  8. 17 Apr, 2004 1 commit