The Class E Network

Zitat von geaves

During the install I picked this up from the log file You are not authorized to access monit. Either you supplied the wrong credentials (e.g. bad password), or your browser doesn't understand how to supply the credentials required.That error is relevant because when I tried to access the web interface it failed with connection refused! So back to monitrc....this was back to the default file and I had to change it again.
I am wondering if the problem is related to the user and password being set to admin / monit as per the tutorial, 'maybe' setting it to the root user and root password used by omv is the answer.

In my case, I simply accessed the monit Web page and noted the potential for setting parameters. I didn't try to change anything. In any case, you'd think the root user would have access.

Zitat von geaves

So there must be a config file somewhere that will allow you to disable monitoring for the remote mount plugin.

Note - this is speculation:

I'm not sure about disabling Remote Mount (monit) monitoring specifically. Remote Mount is a user friendly interface, in the GUI, for setting up a command line with arguments/switches, that is then placed in another config file where the line is activated on boot. (/etc/mtab is used for mounts but it appears to have entries for local devices only.)

(You're probably right, there might a file "somewhere" where it might be possible to turn something very specific like this off, but the question would be finding the file AND the parameter. I've went down rabbit holes in these searches before, tracing through a trail of what seemed like configuration bread crumbs. One config file references the next, go there, repeat...)

In any case, after the mount takes place a "file system" is mounted. Under System, Notifications, in the Notifications Tab, you could uncheck File Systems and, possibly Process Monitoring and see what happens. It wouldn't hurt anything to test it.

Zitat von geaves

Yes, I think if you apply the user root instead of admin and use the root password or leave the admin user in place but set the password as the admin login password to monitrc 'it might work' but it might not because any changes I seem to make that monitrc file reverts back to the default.__________________________________________________________________________________

Yep, the next "on demand" update in OMV (where the yellow "confirm" banner appears), resets a number of Linux package settings to the defaults as known by OMV. This is a strength in that it maintains OMV in a consistent state that can be reported, accurately, in the GUI. For those with more experience, some manually applied changes can be merged into OMV using omv-mkconf. I've used it in one instance, after an edit to /etc/default/openmediavault, for disabling backport kernels.
(actually it was omv-mkconf apt)

Zitat von geaves

Something I've been meaning to ask you, since upgrading to 4 I have not installed any updates unlike 3 when I would update anything that was listed under Update Management.When I was using Mint the kernel was the only thing I set to ignore but updated everything else, but I'm wary of just updating everything on the list....Monday I could shutdown and image the flash drive so at least I have a before option to fall back on.

I'm in agreement with ignoring kernel updates.

And from times past, we seemed to agree that the odds of a software update, damaging a server in some way, are far more likely than a chance encounter with a virus. As you're aware I update "carefully", a few packages at a time, and I really watch updates that affect packages I'm using.
With that said, the only real difference in the way I'm handling updates in OMV4 is by putting a "hold" on the current kernel. (ryecoaaron added this as part of OMV extras in Ver 4. For many reasons, this is an excellent feature for all.)

To prevent issues with my ZFS mirror, when I update my main server to OMV4, I'm going to disable the backports kernel. (But that's another matter). Kernel 4.9.0, the mainline kernel in Debian 9, will be fine for my purposes
_______________________

In any case, I think updating your backup clone is a wise/safe decision before updating the system. An ounce of prevent and all that...

I'm updating USB clones, in prep for the OMV4 upgrade. On the write cycle, what started off looking like it would take about 20 minutes, has ballooned out to well over 3 hours. 16GB drives are starting to look better.

- The clone reading operation, for me, is usually 20 minutes. The write operation is longer but for some reason, this time around, it went way long. (It could have been a virus scan on the client. I didn't check.) It looks like the 2nd drive is going to take an hour.

- On Google searches; I've found them to be more effective than searching the forum itself. If OMV is one of the terms, along with the error, you'll get more hits from a Google search than the built in forum search engine. (If it works, it works.)
__________________________________________________________________________

Yeah, the renovation is in the last stages. The Kitchen is done (actual cabinet making from scratch was involved, modification of store bought items, etc.). What a PITA. Carpentry is one thing but cabinet making requires "dimensional purity" - accuracy to 1/16 is minimum with 1/32 or better preferred. (And I don't have a 20 year old's eyes anymore.)

Now there's a bit of drywall work left, lots of painting, and finally laying down new flooring in about 2/3'rds of the house. The rest, bathroom renovations (detailed but not difficult) and flooring for the extra bedrooms (easy) will wait until after we move in which should be in the next month or two. (Three?) After the move, I'll have to come back to this place to prep it for sale but there's some light at the end of the tunnel.

I still have a few basic structures to build on the other side, a wood shed for example, but I'm looking really forward to a bit of time off this winter. (Or at least to not building anything.) I'm looking forward to the break.

Zitat von geaves

Well done I'll second the eye issue, I have varifocals these are fine for everyday use, but if I get into a tight area where I need to look up I also have to move my head to 'see' out of the bottom of the lense. I now have a pair specifically for computer work, they seem to be a slightly stronger reading lense I had to take distance measurements from my sitting position to the farthest point on the monitors, they work great.

Actually, I have great distance vision that's close to what it was in my 20's, but it starts at 1 yard out. Inside that, like you, I have two different magnification needs. 2 feet (construction measurements and PC screens) and 1 to 1 1/2 feet (for reading). When you mention "varifocals"; I saw a cheap set of readers, in Wal-Mart, that was tempting (about $20US). It had a dial adjustment so one could "dial-in" the magnification needed. (But it seemed way too fragile, for my use. I'm hard on things when in building mode.)

Zitat von geaves

/---/ The wife and daughter are good at painting
/---/ laid wood flooring in the last house
/---/Our kitchen needs a facelift, but the cabinets are good quality, so changing the doors, the furniture and the worktop will be a better option and removing the 'built in' fridge/freezer, much prefer a free standing one far more choice.

- Painting... Rolling it on is easy (but not to be taken for granted). It's the prep work that makes it look good and takes way too much time.
- I'm not doing actual wood flooring. No penny rows in this job. I mean, I want to use it for 30 years or so, not pass it on to my heirs. I'm doing so called "luxury" vinyl plank which is a free floating floor. It looks just like wood but goes down a heck of a lot faster and is easy to repair/replace if damaged.
- My cabinets (the hand built ones) could be refreshed easily later on and, as a bonus, survive a nuclear war. (Picture a blast photo, with the kitchen walls and cabinets still standing.) They have 3/4" material on all sides and 3/4" doors that recess 3/8" into the front openings.

When my wife catches her blog up, I'll PM you with the link. If you skim past her crafts and such, you can see the renovation progress.
_______________________________________________________________

On a PC topic, if you're like me, you're probably supporting family PC's, remotely. Have you tried Team Viewer? It's free for personal use. I was using TightVNC but it required a forwarded port from the router. (With hacks using port scanners, forwarding a port to a Windows box can create a potentially ugly, security issue).

After prompting from a forum user PM:

I looked into the reason why Pi-hole appears to forward IP Ver6 DNS requests. (Without a Ver6 DNS server defined, these entries could be false positives, but I didn't run it down with a sniffer.)

If you're interested, to eliminate Ver6 forwards, the fix is at the bottom of the Pi-Hole How-To.

Zitat von geaves

The requests were all from iPhones and iPads nothing else, but the whitelisting from the Web GUI still does not work, I can add the site, clear the browsers cache but still get the box that the site has detected an ad blocker.

The advertisements themselves (banner boxes, etc.) usually don't originate from the site you're whitelisting. Ad's are generally sourced from outside servers/domains so whitelisting a single web site does not, necessarily, whitelist the advertisements found on the site. If Ad servers or domains themselves are whitelisted, their advertisements would show up on multiple sites.
(Which is why whitelisting an advertising domain is not a good idea.)

It seems that they're adding more features to Pi-Hole so what you're describing, whitelisting advertisements if they show up at a specific web site, may be something that's available in the future.

Zitat von geaves

I would agree, but, I think with pi-hole if you whitelist a site it will display any ad's coming from an external source, so in essence it overrides the ad domain that is being blocked so the ad/s will display on a specific site.

I'll test this when I get back. (We're late going over this week - man, I'm wearing down...)

From your alert - I have the new "pihole/pihole" Docker on my backup server with minor changes to the How-To. So far, so good.
____________________________________________________________

I took a look at OMV 5'S proposed future on github and it seems to be centered on BTRFS.

I had an "experience" with a bit of data corruption on a BTRFS formatted drive which only affected, at most, a handful of files. Surprisingly, it didn't seem possible to isolate the actual names of the affected files. (With that information, I'd have deleted the affected files and replicated clean copies back from the source.)

Correcting the corruption left me wondering exactly what happened. It seemed as if the new checksums, assigned to the corrupted files, where simply accepted as "correct" with other transaction numbers reset as well. As it seemed, BTRFS utilities didn't really fix anything. It seemed to be about reconciling administrative numbers, with the problem itself being ignored.

I don't understand the big deal, regarding BTRFS. I have everything that was mentioned on github, as a future feature, right now with ZFS and it's mature. I have automated, self rotating and purging snapshots, with the easy ability to restore back to any one of several points in time. The only thing BTRFS is supposed to have that ZFS doesn't is the ability to remove a disk from an array and shrink it. (As I understand, that feature has yet to be implemented.)

I must have missed something in the dialog.

Zitat von geaves

That's interesting, because I've been looking at my own setup and wondering what to do, 1) to reduce space 2) reduce power consumption.

Well, I've found that ZFS is not the resource hog many have made it out to be. (Such nonsense is all over the Net.) Without de-dup, which I wouldn't use in any case, it's memory requirements don't seem to be excessive.

FYI:
I'm using ZFS on my backup server, an OMV4 standard build, with 4GB NON-ECC ram and a 64bit Atom processor, and I didn't do anything to ZFS to tweak it for low'ish RAM. It's a power efficient setup, that's a small "cube" in form factor. The data store is roughly 1.4TB. So far it's been fine.
For home users who want to use ZFS with a modest storage requirement (I'll estimate that at 4TB or less), power efficient processors and 4GB RAM seem to be able to do the job.

BTW: You wouldn't need an extra drive for ZFS Snapshots. Snapshots, if used, are an attribute of the pool or child file systems where deleted files and old versions of files are saved. They really don't take up much space at all. (Next to nothing if your data store is relatively static.) If you have a lot of file turn over, in any directory, see this ZFS How-To for the particulars in dealing with that.

Zitat von geaves

For me something simple is all I need, OMV runs, dhcp, docker>>emby>>pi-hole and smb that's really it. I have found another 1Tb drive and a usb at the back of the cupboard and I have been moving old files from my usb 3.0 to that, stuff that I just keep, for those 'just in case' moments

Given that you have all of the older hardware you'd ever need, drives, laptops, etc.; you could probably use UnionFS and SNAPRAID with USB and be fine. I started off with the UnionFS plugin and SNAPRAID, on my backup server, and it worked really well. Frankly, I was surprised. Since SNAPRAID uses file checksums, you'd also get protection from bitrot that's similar to ZFS.
(I tried a ZFS mirror on a lark. Since it worked on weak hardware, I kept it.)

Actually, I'm still using UnionFS on the backup server to merge 2ea 2TB drives into a 4TB Union drive, and using Rsync to do a "drive to drive" copy (the ZFS mirror to the UnionFS drive). That gives me 2 full data backup copies, separated by 1 week in time, on the same box, and it's fully automated. (After I watch this server's stability a bit longer, I'm going to configure automated Snapshots as well.)

Zitat von geaves

I had my first pi-hole complaint yesterday, the wife was out to lunch with some of here old colleagues and daughter was at work.....anyway the wife got back whilst I was just finishing off....and I got 'Oh!! that pi-hole wouldn't let me open a website yesterday' ....to which I replied 'yes dear that's what it's supposed to do, blocks ads on web pages' Didn't like to tell her that it had been running for a quite a few months

My wife is fully on-board with pi-hole and what it does. As creatures of habit, stemming from pre-retirement years, we still do a lot of reading and research. And,, we were finding that nothing breaks one's concentration like video popups, flashing ad's, animated banners, etc. It was becoming a serious problem even on respectable, non E-commerce, web sites. In our case, Pi-hole was joint decision and we're both very happy with it.

Zitat von geaves

1. That's what I'm going to do, have just bought a HP MicroSever Gen7 N54L from that auction site, it means a fresh back up, reinstall and set up./---/
2. the server discovered it's using a 550W PSU this appears to be overkill /---/
3. It would also mean updating the Pi that does rsync..... /---/

1. While I only used it for a couple months, UnionFS (MergerFS) and SNAPRAID made a really positive impression on me. Since I research "stuff", actually using them made me look closer into their capabilities. There's really nothing not to like. The only barrier, I can see, is understanding their concepts and principles.
I wrote a new users guide for OMV (maybe you've seen it) which strongly recommends against RAID with SBC's (R-PI's or other).

However, there are legitimate reasons for aggregating drives under a common mount point such as, for example, having numerous smaller drives. If one wants to aggregate drives and have "RAID like" recovery capabilities; UnionFS and SNAPRAID is the way to go AND they will work reliably with SBC's and USB hubs. (Unlike the known issues surrounding USB hubs and mdadm RAID.) The bonus is the bitrot protection with checksums, provided by SNAPRAID, which makes this approach superior to traditional RAID in SBC applications. The only downside is the loss of the speed boast (parallel read and write) that mdadm RAID provides.

2. It is - overkill. But that's the design criteria for rack mounted servers made "back in the day". For a headless server, 300W would be the most one would ever need. Anything larger than that would result in parasitic losses, even in a well designed PS.

3. Since it has been stone reliable over years, I'm keeping my R-PI / WD 4TB USB drive combo until the 4TB drive fails which may take years... And while the purists will complain about Rsync's down sides, it's worked well enough for me.
_____________________________________________________________

On other stuff - the so-called "luxury" vinyl plank flooring is going down. The kitchen is done, a bit more in the common areas, and I'm working on the master bedroom right now. First impression? Relatively easy to put down (easier, much faster, and far more forgiving than actual hardwood flooring), it's reasonably forgiving of sub-floor imperfections and it looks awesome. Other than looking out of place for a house in a particular price range, you'd have to "eyeball" it to know the difference.

Of course, there's no getting past the trim carpentry. Window trim carpentry and baseboard installation are key to a finished "look". They're both time consuming, and their are no (decent looking) short cuts.