GitLab

multiplex factor (-m). Overrides the setting in the experiments table.

Add -p and -m options to turn on prepass in assign and the mapper, and
also pass the multiplex_factor in.

and neither do I.

Move the important part of os_select into this library as a method so
that we can just do it from the mapper wrapper without having to call
a shell script for each node in the topo (which on a 2000 node
experiment, takes 45 minutes).

beleive we really have this stuff set up to run well at other sites.

Modified the xml format to have attributes which makes the parse much faster. The change affected the rspecs as well and that was also corrected.

page, and include the scripts I used to make it (along with instructions).

things.  Might be the wrong thing to do, but it works...

(we need this for shaping lans).  The imq device relies on netfilter hooks
to accomplish its packet redirection, and netfilter is of course
virtualized.  So, have to capture the movement of IMQ devices between
containers, re-register netfilter hooks on a per-container basis when they
boot/shutdown, etc.  I haven't got rid of the hardcoded limit on imq
devices yet.

shaping, then increment imqidx.

that slows boot on the pc2400s...

regression testing.

speed, 25% faster, or 50% faster then assign_wrapper.

when QUIET was correct <embarrased icon>.

physical state compares properly. Also some actual fixes.

physical state compares properly.

experiment in regression mode, and then compare the results.

wrapper/mapper, so that DB tables are ordered in a consistent manner,
and can be easily compared with just diff.

Also some actual fixes to the mapper, as a result of regression mode
comparisons with the wrapper.

wrapper/mapper, so that DB tables are ordered in a consistent manner,
and can be easily compared with just diff.

mapper wrapper. In regression mode, the wrapper/mapper proceeds
normally, creating a .vtop file, and then running assign with a fixed
seed. If the wrapper and the mapper agree on the .vtop file, then the
solution from assign should be identical.

The wrapper/mapper then proceeds normally, reserving resources and
making all the DB changes. Needless to say, this has to be on a
private copy of the database, with all nodes free. Creating that DB
was a tale in its own right.

At completion, call the existing BackupPhysicalState() function that
we use in swapmod, and write all the physical tables we have changed
(just the rows corresponding to the experiment of course). The delete
all that state, and free the nodes.

If everything is working correctly, those physical tables should be
identical when created by the mapper or the wrapper.

Of course, its not quite there yet. I have a few things to fix up
before diff -r produces no results.

into a function here, which is now called from tbprerun.

In order to run to completion in impotent mode, create new virtual
node objects, but just don't put them into the DB. This lets the
mapper keep going all the way to the end without making any actual DB
changes.