Looking at your barn construction that's what we would refer to as piling, driving a hole into the ground then insert a steel frame and fill with concrete.
At our last house our garden sloped down to the house and right to left, to put in a 12' x 10' 'shed' we would call it we had to dig out footings for concrete then build up in brick to get a level area to 'float' a concrete base for it to sit on. The back footings went down about 4' and the front was around 1' all dug out by hand with a kango (small jack hammer) the spoil taken to a skip by wheel barrow, wouldn't have got an excavator in. I was luck I had a builder friend with a laser level to mark it all out and the depth we needed to go.
The Class E Network
-
Looking at your barn construction that's what we would refer to as piling, driving a hole into the ground then insert a steel frame and fill with concrete.
Actually, those holes were "drilled" with a 12" auger which wallowed the holes out to 14 - 16". The depth was 30", to get below the frost line.
(That's my Massey Ferguson. A tough "anachronistic" 45HP tractor, from back in the day, that can still "get it done".) I could have dug those holes by hand
but it would have taken awhile.The posts are 4X6" with a dovetail wedge cut into the bottom and a lag screw to lock the post into the concrete when poured. With rebar driven into the holes and fencing shaped into a cylinder, the concrete was nicely reinforced.
I nailed up one half of the floor and laid it over 9 pre-drilled holes. I blocked up the floor section, attached the 4x6 posts and leveled it all out. At that point the posts were suspended in each of the holes.
The extra post in the middle, is for a center support on the first floor. If there's a load bearing post above, there should be support below.
(You'd be surprised how often this is missed, even in commercial construction.)The above ground concrete is 12" high. The form is simple, made with 12" squares cut from 3/4" plywood but the continuous length of the concrete column, into the ground, is 42". This is far better than a pile driven into the ground. Seeping into the irregularities of the ground anchors the concrete, and the posts, in place.
(I used a small concrete vibrator to remove all bubbles and to insure that the concrete flowed into cracks and crevasses. This yields much stronger concrete.)With the first floor section in place, posts set, concrete poured and partially cured, etc., it was time for the 2nd section. While a bit more work was involved in this modular approach, work flow had to be designed to be a one man job.
First floor flooring going in. I'm a big believer in Advantec sub-flooring and roof sheathing. (It's also reasonably priced.) It's rated for direct weather exposure for over 200 days. It doesn't absorb moisture (Phenomenal.) Otherwise, trying to cover up with plastic, problems with storms and wind tearing up temporary weather coverings, etc., makes one man construction a real PITA.
_____________________________________________________________________________________________________
@ryecoaaron - I know you're busy with your basement right now but, when you're done:
If you have the land, even if it's on a slope, you can use a similar technique to setup a floor plane and level it out. For a span from the top of a concrete pier to the floor joists that's 2' to 5', I'd use 6"x6" posts. 6"X6" posts are beefy enough. I helped my brother-in-law frame up a house, on a slope, and we used 6X6" posts.
_____________________________________________________________________________________________________And to tie it all into OMV, one must have an offsite structure, that's high and dry, to house your backup device!
-
Webmin and Ubuntu Server - a slow process. While the underlaying CLI programs are the same, the GUI represents a completely different way of doing things.
-
That's what I was using before the switch to OMV, one extra advantage with the above is you can install Cockpit.
You can install Cockpit on Ubuntu and Debian (and others). Debian actually has a newer version than Ubuntu though.
-
You can install OMV on Ubuntu and Debian (and others). Debian actually has a newer version than Ubuntu though
How would that work? OMV4 installs on Debian 9 (stretch). As it seems Ubuntu's numbered versions don't, necessarily, align with a single Debian version.
Would it be a script install? (As in Volker's scripted installs in the Guide page?)
-
How would that work? OMV4 installs on Debian 9 (stretch). As it seems Ubuntu's numbered versions don't, necessarily, align with a single Debian version.
Would it be a script install? (As in Volker's scripted installs in the Guide page?)
Oops. Fixed that. I meant Cockpit can be installed on either.
-
I got OMV5 to install on Debian 9. (I think...
) It says 5.0.1, usul in the Web GUI and apt shows the same on the command line.
If it is OMV "5" it appears to be nearly identical in appearance and function, from a users perspective, to OMV4. (At least at this stage.)Since it appears that the standard file formats are still supported (not a BTRFS only approach), it should be fine. While more than a few plugin cut backs are expected
, I'm guessing that UnionFS, Snapraid, and (hopefully) ZFS will still be supported. If not, setting them up on the command line is doable, so,, I'm almost relieved.Here's to hoping nothing changes radically as OMV5 is developed.
-
I got OMV5 to install on Debian 9. (I think... )
Working fine here too.
If it is OMV "5" it appears to be nearly identical in appearance and function, from a users perspective, to OMV4. (At least at this stage.)
Yep. All changes are backend changes (configs created by salt).
While more than a few plugin cut backs are expected , I'm guessing that UnionFS, Snapraid, and (hopefully) ZFS will still be supported.
Here is the list I am maintaining - Why should you use or not use dockers
-
Here is the list I am maintaining - Why should you use or not use dockers
That list seems to be both logical and reasonable. Most of what you're porting to 5, in the way of plugin's, would need direct integration into the OS. (To include one I know you dislike, PHP Virtualbox.) The remainder make sense running as Dockers because they would run well as an independent server.
Now if we could only get Volker to give Docker support it's own section on the forum... That would segregate a lot of Docker related questions (that we know are coming) away from OMV. They're two distinctly different topics.
_________________________________________________________________Part of what I read on Volker's ideas (Git Hub), on going forward, had me worried. I can't find the exact reference but it was something about focusing on BTRFS and it's sub-volumes (I imagine BTRFS sub-volumes are the equivalent of ZFS filesystems). I've had enough experience with BTRFS to know that parent transid verification failures would be a problem and it only takes a handful of events to trigger an issue that can't be recovered.
Given the environment I'm in now, I chose BTRFS for it's "Copy on Write" abilities where nothing should be lost in the event of a power outage and low hardware requirements. The idea was to use a filesystem that is "forgiving", in a portable application (no UPS). In a relatively short period of time, I had two separate events that were not repairable, where it was necessary to restore from backup. So, while data may not be lost in the literal sense, if the filesystem fails due to corrupted internal housekeeping processes, the net effect is the same - data is lost. (In roughly the same conditions, EXT4 had no problems at all.)
While Dev's run hosts on an UPS, I'm of the opinion that a significant percentage of users do not. With numerous users running without an UPS (potentially numbering in the 1000's), my "Spidey Sense" tells me that focusing OMV development on BTRFS would be a disaster. It simply is not ready yet.
-
(To include one I know you dislike, PHP Virtualbox.
I left virtualbox on the list even though I want to just say use cockpit/kvm.
can't find the exact reference but it was something about focusing on BTRFS and it's sub-volumes
-
Here, at the new place, one could set up a power line effects testing lab. (One month here would equate to six months or more in most other areas.)
I don't know what the thresholds are for my UPS (dips, spikes, under/overvolt, etc.). It might be overly sensitive. However, a few power events per day are not uncommon. I'm going to plan on swapping out UPS batteries every 18 months to 2 years. (They usually last 4 years or so.)
I'm glad I installed a panel mounted, whole house, surge suppressor when I was doing the electrical work. If the line behaves badly, it will shield our electronics and appliances.
_________________________________________In the bottom line, I believe BTRFS doesn't handle power outages gracefully. In similar conditions, the simple journal approach EXT4 uses recovers every time. And while power may be stable in your area, it's not as stable in other regions, even among some of the more advanced countries in the EU.
Given current events, it doesn't matter in the short term. However, if BTRFS becomes the standard file system AND it doesn't improve significantly, I believe the support issues flood gate will open.
-
I believe BTRFS doesn't handle power outages gracefully.
XFS doesn't handle power outages all that well either.
-
We'll just have to wait and see re BTRFS becoming standard, just don't upgrade, or move back to Ubuntu or Debian. I'm of the opinion 'if it ain't broke don't fix it'
It isn't necessarily about upgrading. btrfs allows you to do things that other filesystems can't. Commercial NASes are offering these features and typically use zfs or btrfs to do it. It is difficult to support both filesystems without a plugin and I think Volker would like to focus on the new direction. We should be able to write a plugin to support the old way.
-
btrfs allows you to do things that other filesystems can't
That I like as I used zfs on what was nas4free, but eventually I gave up on it because for whatever reason it would throw a hard drive, the same hard drive every three months, only for me to remove it test it, run dban then add it back. Changing ports, cables made no difference it would error again, so I gave up and moved to Ubuntu a few months after that I found omv.
TBH, I think it's Flmaxey's power outage reports that's putting me off, but we don't have that problem in the uk.
-
Well, it appears all is good for OMV5, if the beta ships in a state that is close to what it is now. If the latest version of Debian (Buster) is used in the final 5.0 product, it should be good for at least the next few years.
if it ain't broke don't fix it
That's the mechanic's or technician's first commandment. (And the 2nd, 3rd,
.)
___________________________________________________________Apparently, there's some hub-bub over ZFS (ZOL) not being supported in Linux Kernel 5. And it even appears that Linus Trovalds is expressing some annoyance regarding the idea of coding the new kernel for compatibility.
This could become "interesting". Can a Kernel be forked? Or should I say, does a bear crap in the woods? I can answer the latter for sure; "Yes", they do!
__________________________________________________________I've been poking around in the Webmin interface, in Samba Windows Networking. It seems that the latest selectable protocol level is NT1? That goes back to the WinNT days. I'm fairly certain that Ubuntu server's Samba package is good to SMB3.0
I must be missing something.
-
if the beta ships in a state that is close to what it is now.
It will keep changing. Trust me. That is why I haven't ported all of the plugins.
If the latest version of Debian (Buster) is used in the final 5.0 product, it should be good for at least the next few years.
It will use buster. And Volker won't release it before buster 10.1, I'm guessing. So lots of time for changes.
Apparently, there's some hub-bub over ZFS (ZOL) not being supported in Linux Kernel 5. And it even appears that Linus Trovalds is expressing some annoyance regarding the idea of coding the new kernel for compatibility.
ZFS uses kernel features that have been deprecated for a decade!! So, it isn't like they didn't know it was coming. Someone will fix it.
Can a Kernel be forked?
Yes but it won't happen. Too much big money on the kernel. A fork would make a big mess.
I've been poking around in the Webmin interface, in Samba Windows Networking. It seems that the latest selectable protocol level is NT1? That goes back to the WinNT days. I'm fairly certain that Ubuntu server's Samba package is good to SMB3.0
I must be missing something.
webmin probably hasn't been updated. Ubuntu and Debian 9 support SMB3. Just need samba 4.5 or newer.
Jetzt mitmachen!
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!