Topics created by bbigford

You were lucky enough to move there in an interesting year. Boise saw record snowfall last week, and it is true that a city only plans for 'the norm'. When planning for the norm proves ineffective, only then will they invest in new plows. Every city has posted evacuation routes in case of emergencies. Those evacuation routes (often residential side-streets that are otherwise not often used) are always the first to be plowed. Then they move to main arteries, then busy residential arteries. In a bad storm, most other residential streets are ignored entirely.

I'm in Montana. Our past 5 winters have been some of the worst on record. The first one was a nightmare; nothing was plowed. The plows the city did have were being run 24 hours a day on shifts, they were breaking down, etc... the breaking point was that they couldn't keep up with clearing evac routes. Over the next few years, our city has gotten VERY good at plowing. We have some seriously behemoth machines. Plows travel in pairs so they can clear an entire street in one pass, they usually run from 9 PM to 6 AM to not interrupt traffic and only spot check during the day. You almost never see them and the streets are always clear. It's amazing how far they've come. This winter has been nuts so far. Shovel 12" off the drive, shovel 10" the next day, then the wind picks up and drifts, so you shovel another 12" of wind packed ice. Rinse repeat for 3 weeks. My brother has been running his 4-wheeler up and down streets charging people to clear driveways.

tl;dr: They're right. If Boise rarely sees over 4", it's hard to justify new equipment in the budget. Buy yourself a 4-wheeler with a plow and charge $50 to clear driveways. It'll pay for itself in a week.

@BBigford said in SQL understanding - power outage:

@scottalanmiller said in SQL understanding - power outage:

@BBigford said in SQL understanding - power outage:

Here's what the person I heard it from said:

"likely easy to talk on the phone (i'm on a conference call, but have time this afternoon).

The term is quiescence, that you are looking for. This means that writes have been halted, and all data in buffer has been flushed. Part of the challenge here is this needs to be done not only at the app layer, the OS layer, the storage layer underneath that (and even at the drive level, as cheap consumer SSDs and SATA drives will have write buffers not protected).

Modern SQL commits writes first to the transaction log then to the database, and in the event of a power loss can "replay" the transaction log. The problem is if the log is large the can take a REALLY long time, and if you have systems underneath that ACK'd writes at the SCSI layer out of order (Consumer level SSDs) this might not properly recover."

Not really sure what data is sitting in the buffer during a power outage though. Meta data waiting to be written to tran logs? So then when a power outage happens and that buffer gets flushed, there's an inconsistency in those tran logs that were mid-write...

That's where atomic commits come in. They make sure that everything is quiesced before anything goes to the database. ZFS does the same thing for different tasks. This really isn't an issue in a properly designed database - unless you've added a caching layer yourself that doesn't honour the flush commands from higher up the stack. Then it is on your own head. You could always put your database into RAM and cause it to fail like that if you wanted.

I found out what he was talking about. Commit to tran log > Application acknowledge transaction > written to DB. The data that gets flushed shouldn't matter because most applications wouldn't acknowledge something that isn't commit. So if there was a power outage then the tran logs could just replay, app acknowledges, and so on.

I was missing something, and only thinking of tran logs and the database. Didn't think he had been talking about the front end.

Ah, okay. Yeah, if the application doesn't wait for the database, it's just being reckless.

Congratulations and good luck on the new position.

Not sure a non admin can uninstall UWP apps. Store access should be available to admins.

@BBigford said in Hyper-V with Smart Card not working:

@scottalanmiller said in Hyper-V with Smart Card not working:

@BBigford said in Hyper-V with Smart Card not working:

@scottalanmiller said in Hyper-V with Smart Card not working:

@BBigford said in Hyper-V with Smart Card not working:

@scottalanmiller said in Hyper-V with Smart Card not working:

@BBigford said in Hyper-V with Smart Card not working:

@scottalanmiller said in Hyper-V with Smart Card not working:

@BBigford said in Hyper-V with Smart Card not working:

saves from pasting a lengthy document, plus a few images.

Posting images is fast and easy. And what takes five seconds for you to do, saves each reader of the page many seconds for forever.

The re-formatting of 5 pages into copy-paste took longer than just attaching the document. I did a straight copy-paste in the OC.. haha it was real bad.

I know it takes a bit if you are coming from a huge rich Word doc. But it is what it is, Word docs are definitely not something we are comfortable hosting. That puts ML on the hook for malware hosting. It's a huge deal. The whole site could get blocked in a split second because one newbie doesn't scan a document that they upload.

If it were only PDFs, it could be considered. But there are so many ways to host and link those things today that I'm not sure that it is a big deal. You can link it when really necessary and, whenever possible, it's best to get text and/or images instead.

It is what it is. I added stuff to the OC and will spend some time formatting the copy-paste. There's some good stuff in there for readers, and can explain what is going on and what I'm seeing... I thought it would be helpful.

If you output to text from Word, does that help?

I'm doing this from a Chromebook now and just did it within Google Docs. Doesn't have the ability to export our mess with output since it's a pretty basic tool. I'll have to try that a bit later when I jump on my laptop. I added the images.

Check out this tool:

https://word-to-markdown.herokuapp.com/

Completely locks up on a Chromebook trying to paste. Switched to a Windows 10 workstation with some more horsepower, all rendering/copy/paste goes through quickly. Copy to Clipboard appears to not work at all. Formatting looks terrible. But I'm done spending more time on that.

Well that sucks, it looked interesting.

Lol, sure... But if a graphics design person can't work out this situation, they probably belong on another line of work. Considering the OP. 😉

@jospoortvliet said in NextCloud box:

@Reid-Cooper said in NextCloud box:

It's a hobby device. NextCoud said so themselves. Not meant for business use. Just for fun.

Yeah, certainly not a business device, it isn't powerful enough for that. It is meant for home users.

And yes, it is designed to not require maintenance. That is, the OS automatically updates, as does Nextcloud itself. I think that at the moment, you'll have to re-enable apps, that is the only thing. We're working on making that work automatically too. Help is obviously welcome.

Oh OK this just got a lot more interesting. I personally feel that anything that goes to a consumer needs to handle 99.9% of maintenance tasks completely on its own. If it doesn't then most of us know that in a consumer environment it will never happen, and we end up with IOT devices helping to cause 1 TB DDOS attacks.

@wrx7m said in Offline files nightmare:

I had a freezing issue where the system would just stop responding for like 30 seconds at a time. It wasn't the well-documented one because I have a single partition for the OS and data. There was also another issue with the way that it was handling multiple networks being connected. After reverting, it worked fine. I tried installing it several weeks later with all patches and it seems good to go for now.

Awww, I did hear about that issue, just never saw it. All of my systems have the extra partitions as part of install.

Haha...

http://www.boxtech.com/

https://technet.microsoft.com/en-us/library/ff730963.aspx

The door to door salesman normally don't even work for the ISP they work in conjunction with the ISP in order to sell contracted services at higher rates. These guys work purely off commission in most cases so they're pretty good liars as well. CenturyLink is the old qwest... they weren't good either. It's really shitty service here in Denver with abnormally high downtime and poor speeds. They do the same thing the cable providers do and over stuff trunks.

@BBigford said in Thumbtack - what a ripoff:

You're basically paying $3 every time you quote someone, to which you may never see a return.

If a client is sitting on 10 quotes, Thumbtack makes $30 on that quote.

There are times where I don't have to pay at all to communicate with a customer due to a lack of pros in their area.

@prcssupport said in That moment when...:

I could take this job and cause the system to heat up so much and kill the bugs remotely... would I then be a remote bug exterminator

You probably just need an IoT air compressor. As we all know, IoT solves everything 😉

@BBigford back on the original question though. You never answered a damned thing.

Forget AWS/Azure.

What services does this friend want to move? Answer this and we can give intelligent on topic answers.

@BBigford said in Dell R710 NIC teaming issues:

@JaredBusch said in Dell R710 NIC teaming issues:

@BBigford assuming you had no team made, your screen looked like mine?

When you then clicked on team at the top did you have create? Because I do not.

Yeah, Create is now gray but only cause the NICs have all been allocated. Before, it was the only option in the menu. Pretty different from BACS 3, and not sure if that's a good thing... 😐

0_1474655858474_3.jpg

I do not get that option. So yeah, it must know the NICS are not available even though it is not managing a team.

@BBigford said in Serious question about Linux security...:

@scottalanmiller said in Serious question about Linux security...:

@BBigford said in Serious question about Linux security...:

@scottalanmiller said in Serious question about Linux security...:

@BBigford said in Serious question about Linux security...:

I've been under the impression that's how it always was... Red Hat patches a vulnerability or changes something, then CentOS does.

It's how it was when CentOS wasn't part of Red Hat. Now CentOS isn't a company, just a product of Red Hat. So Red Hat is patching both. So your statement above can be rephrased to...

I was under the wrong premise. Whoops... thanks for the clarification. Did not get those answers over at SW. Another reason I love ML.

http://www.linux-magazine.com/Online/News/Red-Hat-Adopts-CentOS

Lots of big time CentOS users here 🙂 It was good before RH bought them, but way better since they did. Used to be even months behind in releases. Not like that at all anymore.

RH has to keep CentOS patched to maintain their reputation as the most enterprise OS option in the SMB and commodity spaces.

Yeah I live under a rock. We don't get updates about that kind of stuff in the desert. Also, please send food and Internet.

0_1474563045734_rock.jpg

+1 internet sent.

@thwr said in Seriously, thank you.:

@JaredBusch said in Seriously, thank you.:

@thwr said in Seriously, thank you.:

@JaredBusch said in Seriously, thank you.:

@BBigford said in Seriously, thank you.:

The threads here seem to start with a focus, and when the issue is resolved, it often spins off into another direction of good conversation. Not always a bad thing since the conversation can continue and the OP can just stop watching if need. Where other sites discourage that, which doesn't make sense cause you really don't need to start a new topic for every conversation shift. But I digress.

It is better that things get forked than continue in the original thread if they diverge too far. Otherwise, it makes it impossible to tag and find later.

One thing that makes ML special compared to other communities is exactly that. Problem solved, but the discussion may go on often talking about other aspects that are somehow related to the OPs question. @JaredBusch is right, this is a problem due to (additional) content not being tagged and therefore hard to find in the future. On the other hand, I think that this is what drives the community.

The kind people of the @Mods group started to fork threads lately when there is an interesting discussion that may be good to keep around, correctly tagged.

Right, additional organic conversation is great. Just once it diverges enough it is better forked just for the tagging and searchability.

If NodeBB would add tagging for individual posts in a thread that would be great.

A "vote-for-fork" feature could also be a good idea.

A vote for fork is vote away from spoon! #dontbesheeple

@gjacobse said in SMB firewall options:

@zuphzuph said in SMB firewall options:

@BBigford said in SMB firewall options:

@zuphzuph said in SMB firewall options:

Untangle. 😄

You've gotten to mess with that more than I have. Have you checked out the content filtering and such? Does it have a VPN client? I couldn't remember if OpenVPN is available on that or if I'm thinking of pfSense...

https://wiki.untangle.com/index.php/IPsec_VPN

Don't feel bad @zuphzuph - I installed Untangled at two NPOs and and if I could have found a small enough PC to run it at home years ago I would have. However - I was 'coerced' into not doing so and that it was a bad idea.

I have a UBNT ERL here at home now, I know one Untangled box I installed has been pulled, as it's now in the scrap pile here at home. Since it was hardware I had bought personally, they gave it back to me (minus the drives).

There are times though I still would like to have one running, just for giggles..

Feel bad about what? Find a bridge.