GitLab

This reverts commit e709749894d85768044ce0631588385ad06ac06b.
Don't need it after figuring out imagezip

Better for VMs where there might not be NFS access (as in Utah),
and because in general, using NFS is bad, use the web server
instead.

This completes linkdelay support for linux, dropping IMQ in favor
of IFB and netem (so that we do not need to build a special kernel).

Have sdcollectd log the IP address associated with the incoming packet when
it can't map it to a node id.

Since the pcpg-i2 nodes are so very flaky, lets try something that
does not require them to rebooted or imaged!

The key change is that on these remote nodes, we do not bridge the
physical control interface to the VM control interfaces. There is no
point since there are no routable IPs we can use, nor is there a
192.168 network that would be useful.

However, we still want to give the VMs their 192.168 address and we
still want multiple VMs on the same host to talk to each other, and we
we still want the VMs to be able to access the outside world with NAT.
So we still create the xenbr0 bridge, and give it the router address
(192.168.0.1). Any traffic heading out will be NAT's as normal, and
you can ssh into the VM using the physical host IP and the per VM
sshd port number.

* remove check for gather arg, use TESTINFO from tmcc hwinfo
  instead. Only call tmcc once. Let non-MFS boot collect data.
* tmcc hwinfo not return a TESTINFO line. Handle the LOGDIR, COLLECT
  and CHECK indicators
* Do to variable scoping make sure everything runs in the shell.
  Do not 'exit' from called programs. Source called programs instead
  of starting new shells.
* Nothing wrong with the which() function I put in since the
  MFS/busybox installation does not have one, but if we have a real
  which(1) program use it.
* output format changes and programming space changes
* remove check for gather arg, use TESTINFO from tmcc hwinfo
  instead. Only call tmcc once. Let non-MFS boot collect data.
* tmcc hwinfo not returning a TESTINFO line. Handle the LOGDIR, COLLECT
  and CHECK indicators
* Do to variable scoping make sure everything runs in the shell.
* Do not 'exit' from called programs. Source called programs instead of
  starting new shells.
* Nothing wrong with the which() function I put in since the MFS/busybox
  installation does not have one, but if we have a real which(1) program
  use it.
* call sub-programs by sourcing not by starting new shell.
* oops mfsmode is not set until initialize is run. also set -e -u to
  catch this type of thing.

Checknode tests once again run in standard images and can collect
data.

* All tests can now be "sourced" from gatherinv to insure everything
  runs in a single shell.
* All tests can also be called standalone.
* MFS mode and non-MFS mode honor the CHECK and COLLECT flags.
* Create, if necessary, a new project save directory.
* This will be needed once every-time checknodes is run in a new project.
  path is /proj/{pid}/nodecheck.
* If script can't run quit with a 'return 0' so the rest of the checks
  can continue.

Change filename gatherinv -> checknode

* Can't run in MFS mode if not in the emulab-ops pid

This is the client-side of our current "swapmod" goal. You can add and
remove blockstores, but you cannot modify the attributes of an existing
blockstore.

Prior to this commit, XEN guests disks were single partition, no MBR,
the bits dumped into the lvm. This makes a snapshot of a XEN node,
look completely different then a physical disk image, especially if
users want more disk space (mkextrafs) inside the guest, and then want
to take a snapshot of that, and then run it on a physical node (which
was not possible).

With these changes, guests now use the same MBR layout as our version
two MBR, which makes them interchangeable with physical disk images.
In fact, the goal is to be able to switch back and forth as needed,
based on physical resource availability.

We were assuming that if static_widearea_config returned 0, it had
configured the physical $iface that findcnet was passed successfully.
But, static_widearea_config also returns 0 if the widearea config info
was present BUT if the given physical $iface did not match the widearea
config info.

So, catch that case, and all is well.

Also, fix a minor syntax bug; and add some debugging.

Part of our hack strategy to detect and correct transient NFS errors
is to fsync after every file write and check for errors. However, some
devices don't support fsync. So, for example, if you were trying to
imagezip to /dev/null (to benchmark or check the resulting image size)
fsync would always fail and we would keep retrying the write forever.

There is already a "-N" option to avoid such checks and retries, but
Mike can never remember that!

The script still blows up on a glob of an empty directory.

What source to read the info from and which enternal array to fill.