GitLab

When destroying a dataset it is necessary to remove all snapshots,
even those not part of the blockstore subsystem. This adds a path
to allow that.

We were still trying to allocate boss/ops nodes with FBSD83 and
inner nodes of type pc600 or pc850.

This way we can create backup snapshots of all datasets without
causing any problems with the blockstore code.

Some messages said bytes when it is really bits, sizes were actually
Mib/Gib instead of Mb/Gb.

No doubt all hell will break loose since "All I did was change a comment!"

Whenever I invoke $image->hash() to get the SHA-1 hash of the image file,
I instead get back a ref to the internal Image.pm $self->{'HASH'} hash
where we store fields from the DB. At least this is the case on FreeBSD 10.3
(perl 5.20). Some odd interation with the AUTOLOAD function?

Anyway, just use $image->field('hash') instead.

Part of an effort to make delta images more precise. But the increased size
of the images means this probably isn't worth it. Revisit when I have more
time (ha, ha!)

We do not appear to go through this path very often!

1. For multi-architecture, being fixed to a specific node is the same as
   being fixed to a specific type.

2. Watch for return value from nalloc, where -1 becomes 255. Hate that.

Also a little repourl cleanup.

of these now, time to start managing them properly (we have lots of stale
ones in the DB).

last revision. Such a dope.

(By assuming nobody swaps in, extends, voluntarily swaps out, or requests
reservations.)

(There is room for improvement here... in some cases we could and
should still perform partial allocation.)

The control flow is kinda confusing, cause of the desire to hold off on
converting the geni-lib script (reducing the time we are updating from the
origin repo in the first please wait modal). Not really happy with how it
turned out.

I also need to change the code that does the initial fetch of a new repo,
and then a delete. The only reason for this is so we do not leave dangling
repos on our file system if the user wanders away before submitting the
form. Need a better way to deal with this, many options available, but
probably need to do something soon since the overhead (second fetch) on
larger repos will be annoying.

mebi (using Blockstore.pm routine) and do not permit repositories over
5MiB for now. This is clearly temporary.

reboot/reload/snapshot, except when the instance is locked down. Add
additional test for project leader in this case.

to prevent messing with speed/duplex on a per-port basis.