@anthonyh said in SpamAssassin Question:

quick reply-of-shame, then an update of this thread.

Nice! Is like reading your code in a nice screen 🙂 just for what you need

@JaredBusch Could be. All I know is it's working now.

Great guide! Super helpful.

@scottalanmiller said in BackBlaze B2 beating those prices down. . my god:

@Dashrender said in BackBlaze B2 beating those prices down. . my god:

@scottalanmiller said in BackBlaze B2 beating those prices down. . my god:

d pretty quickly you've got a sub-$10K server being hosted at $100/mo. That's a $500/mo skew.

The time value of money and your time managing the device are the next things to consider in this cost equation.

If you check the math, I included some in there already. Maybe too much, maybe not enough, but some is there.

Missed your new post before making mine.

ok, so according to the presentation a netscaler is:

a load balancer a reverse proxy a SSO gateway a (W)AF (web application firewall)

now first two points can be obtained by haproxy/ngnix

point 3 can be obtained -just googled- for web apps by naxsi + nginx or modsec + nginx/apache.

also point 3 for non web apps can maybe obtained via these softwares (wiki)

I miss point 2 an - with a bit of effort - I have my free open netscaler 😛

Configuring MPIO to automatically claim all iSCSI devices

Enable-MSDSMAutomaticClaim -BusType iSCSI
0_1494445991062_mpio-iscsi.png

@EddieJennings said in Hosts file and DNS:

On centrally managing hosts file, is there some kind of tool you use for that, or is it literally, just have one hosts file stored somewhere, and copy it to X machines?

Just having something like your Jump box that already has permissions to access each machine and just having a script that automatically puts it on each box is definitely one way to go and super simple.

Salt, Ansible, Chef, Puppet, cfEngine are all specifically built for this, too.

Zoho is definitely pretty nice, and if you fit into their free tiers it's free, which is a pretty big deal.

@scottalanmiller Done 🙂

@Grey said in Printer clustering?:

@scottalanmiller said in Printer clustering?:

@Grey said in Printer clustering?:

Yes, I'm aware of the load balancing. It's really neat! I wish this were so simple. I just need fail tolerance, but your idea of location based printing is a great idea, too.

DFS is not like fault tolerance, I think that's the confusion. If you just want HA print serving, use CUPS as the print server and make it HA however you like. Normal OS HA, VM HA, whatever.

Not using Linux.

Well just fix that, then. Why not?

@Dashrender said in DNS discussion:

@scottalanmiller said in DNS discussion:

@Dashrender said in DNS discussion:

@scottalanmiller said in DNS discussion:

@wirestyle22 said in DNS discussion:

@scottalanmiller said in DNS discussion:

@wirestyle22 said in DNS discussion:

@scottalanmiller said in DNS discussion:

@wirestyle22 said in DNS discussion:

@scottalanmiller said in DNS discussion:

@wirestyle22 said in DNS discussion:

@scottalanmiller said in DNS discussion:

Who DOES determine that you will not fix things properly today? Who is making that decision right now and what was their reasoning?

I've realized that if I wanted everything to be done the way I think they should be done I need to create my own MSP, which I'm not ready to do yet.

No, MSPs make no decisions. Customers make those decisions. If you want to do things the way that you want them done, you have to stop working at an MSP and start being a business owner. There is no other position that gets to do that.

In this case I'm a subcontractor so I have 3 levels of people to convince:

My boss The company that holds the actual contract with the city. The city themselves

If I were an MSP I'd eliminate two of those

You still only have one person to convince, the city.

It's an improvement 😄

You always have the customer to convince

No, I mean RIGHT NOW. You , today, have only the city to convince. Those MSPs above you are not decision makers.

I don't interface directly with city management. The company who holds the contracts do so if they decide to not do something it would never be brought up to the city in the first place

You are the sole on site IT advisor, and they don't talk to you? They literally have on interface between the CIO and the senior IT person?

Hold the phone Scott - You even told me that there is NO REASON that the city in a case like this should ever talk to the onsite tech. Because talking to the tech could lead to all sorts of miscommunications. I.e. the MSP has other plans the tech might not be aware if, etc, etc, etc.

and now you're asking if the customer has a direct line to talk to an onsite tech? WTH?

Ah, but he's not the "onsite tech", he's the senior most tech for all of the associated organizations, including the city itself. That he is on site is not the relevant piece. That's his juniors and you are right, I don't feel that the CIO should be talking to them. But this is the CIO or whoever is running IT refusing to talk to the only possible technical interface point. That's the issue, not that he is on site.

The city needs to either...

Remove their role in IT management or... Have IT management interface with the IT technical people.

Oh, then I think Wire might have lead you down the wrong understanding. I don't think Wire is the most senior IT whatever - his boss is, or probably more accurately, the MSP who actually holds the contract with the city, is.

Or, I completely misunderstand his role 😉

His boss is the BOSS, but not on his technical level. His boss is an L0 or more likely L1 from the descriptions. He's an extremely junior tech on the site.

CalDav for Exchange... http://davmail.sourceforge.net/

@travisdh1 said in Dell VxRail:

@scottalanmiller said in Dell VxRail:

@NetworkNerd said in Dell VxRail:

There's an entire Ready Node selector here on VMware' site - http://vsanreadynode.vmware.com/RN/RN. One thing we learned is that you can go single-socket with the hosts to save on vSphere and vSAN licenses. Just beef up memory because there will be some overhead for vSAN traffic that folks seem to overlook. I know we did.

Most vendors are pushing single socket these days, except for MS. That's why Scale starts on single socket, for example.

That actually makes sense. With got 22core/44thread processors available, the need for more than a single processor is far less than it used to be.

Yes, it's a very common SMB mistake to buy two procs when almost never is that warranted in the SMB market. Single proc of the same core count and speed is faster than two procs and cheaper.

@hobbit666 no this is just for the mysql database.

@scottalanmiller I did not get to mine. I got sidetracked by Kemp.

@NetworkNerd said in VMware Essentials Info?:

... it's time to go somewhere at which technology is perceived as providing value to the business.

Same as saying "where they perceive the business as having value."

@NerdyDad said in nslookup - can it use WINS info:

@scottalanmiller Just out of curiosity, is there an nslookup alternative command for WINS? If so, what is it?

nblookup.exe

The Scale partner solution that people are thinking of is Work Spot. And no, it does not automatically create the VMs it's more of an access layer.

@Dashrender said in Cameyo - Application Virtualization:

@RojoLoco said in Cameyo - Application Virtualization:

Didn't @art_of_shred recently get a Cameyo?

I don't know about recently 😛

Maybe at MangoCon it can make a Cameyo appearance!

@Tim_G said in SAM's Chromebook Adventure 2017:

I'm thinking of getting a Chromebook just so I can have non-mobile (desktop mode) web access in places I do not want to have a normal sized laptop. Is 11.6" too big for that purpose?

Phone is just too small and so inconvenient for some things I want to do. I'd rather have a super small laptop just for web access and simple things.

11.6" is not very big. It's only slightly different than carrying around an iPad, but for me, way more usable.