Yes, what FreeNAS calls a Volume, ZFS calls a pool. It's possible to have multiple pools, but you (and most users) only have one. So, yes, if you pull the wrong drive, you shouldn't be able to access any of your data.
Your configuration right now has two disks (call them 1 and 2) mirrored, striped with two more (call them 3 and 4). Right now, we'll say that #2 has failed. Since it's mirrored with #1, your data remains intact. Once you replace #2, it can fail again, or #1 can fail, without harming your data. However, if #3 or #4 fails, your data is gone forever. I'd guess you initially built this server with two disks, and later added another disk, and later yet another, as you needed more space.
The best answer on pool layout (and a lot of other things you should know) is
@cyberjock's powerpoint at
https://forums.freenas.org/index.ph...ning-vdev-zpool-zil-and-l2arc-for-noobs.7775/. In brief, though, your same four disks in RAIDZ1 would give you the same capacity you have now, but tolerate the loss of any one of those disks. That is, any single disk could fail without harm to your data. A better solution would be to use RAIDZ2, probably with more or larger disks. RAIDZ2 will tolerate the loss of any two disks. It will also protect against a read error when resilvering after the loss of one disk. But, of course, it takes more space for parity.
I do think one of your disks is failing, or has already failed. If your server has booted with ada3 disconnected, but the pool isn't available, then ada3 was the wrong one. There's a better way to definitely identify which one it is, which I'd forgotten when I last posted. It takes a couple of steps at the shell.
First, run 'glabel status'. That will return a list of gptid numbers (long numbers), along with the partitions they refer to. Find the disk that corresponds to "gptid/7ace49e7-733e-11e3-ae8b-d850e6db3a84". This is the disk in your pool that's failed. It will say something like "ada2p2", but you aren't interested in the "p2" part.
Second, do 'smartctl -a /dev/adaN | more', where 'adaN' is the disk number that's returned from glabel status. As before, note the serial number, power down the server, plug in the disk you previously unplugged, unplug the disk that matches the serial number, and reboot. If your pool is now available, you'll know you unplugged the right one, and can proceed with the replacement.