This person was running on-premises Exchange, which you aren't doing, but the principles still apply.
Use redundancy and virtualization. You will thank yourself later.
Two physical servers, each running a virtualization platform. (Either VMware ESXi free version or Microsoft Hyper-V work, and both are affordable.)
One Active Directory server on each physical box. (Each Active Directory server will also be a DNS server. It's usually advisable you use Microsoft DHCP as well.)
My recommendation if you're limited on server hardware is to use Hyper-V. I've done this in small environments before and it works fine. An example might look something like this:
First Physical Server: HYPERV01
VM: ADS01
VM: EXCHANGE01
VM: FILE01
Second Physical Server: HYPERV02
VM: ADS02
VM: EXCHANGE02
VM: FILE02
Total cost: 6 licenses for Microsoft Server, 2 licenses for Exchange, 2 physical servers, plus however many CALs you need. You can do this for around $10k before CALs.
If you can't quite swing that, you can also do:
First Physical Server: HYPERV01
VM: ADS01
VM: EXCHANGE01
Second Physical Server: HYPERV02
VM: ADS02
VM: FILE02
You don't get redundancy on Exchange or file sharing, but you can easily migrate those VMs to another physical server if something crashes.
Your software cost will be a little less because you're not doing Exchange. But you will also need to think about a rack or cage, UPS and power distribution, possibly cooling.
Feral on
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
the "no true scotch man" fallacy.
0
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
This person was running on-premises Exchange, which you aren't doing, but the principles still apply.
Use redundancy and virtualization. You will thank yourself later.
Two physical servers, each running a virtualization platform. (Either VMware ESXi free version or Microsoft Hyper-V work, and both are affordable.)
One Active Directory server on each physical box. (Each Active Directory server will also be a DNS server. It's usually advisable you use Microsoft DHCP as well.)
My recommendation if you're limited on server hardware is to use Hyper-V. I've done this in small environments before and it works fine. An example might look something like this:
First Physical Server: HYPERV01
VM: ADS01
VM: EXCHANGE01
VM: FILE01
Second Physical Server: HYPERV02
VM: ADS02
VM: EXCHANGE02
VM: FILE02
Total cost: 6 licenses for Microsoft Server, 2 licenses for Exchange, 2 physical servers, plus however many CALs you need. You can do this for around $10k before CALs.
If you can't quite swing that, you can also do:
First Physical Server: HYPERV01
VM: ADS01
VM: EXCHANGE01
Second Physical Server: HYPERV02
VM: ADS02
VM: FILE02
You don't get redundancy on Exchange or file sharing, but you can easily migrate those VMs to another physical server if something crashes.
Your software cost will be a little less because you're not doing Exchange. But you will also need to think about a rack or cage, UPS and power distribution, possibly cooling.
If money is a SERIOUS issue, one physical server is an option. Not the greatest option, but budget constraints are real.
It's not that they can't afford either, it's just owners are cheap asses and balk at $20,000 price tags.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
0
Options
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
I will say that if the company is only willing to spend about $5k on an Active Directory implementation, I’d consider getting one physical box with good redundancy (good RAID, dual power supplies, etc) in its own hardware rather than trying to squeeze a bunch of VMs on there just for the sake of virtualization.
I’d also question an installer who wants to do that for $5k. In my opinion, to kind of go along with what Feral was saying, if you’re getting into the virtualization game, I think you should be at two hosts and a NAS to begin with if you want to do it right. I haven’t specced out low cost VMware hosts in a while but I think you’d be in the $20k range total, maybe less if you compromise in a couple areas. But I do have to admit that my virtualization budget is rather good compared to a lot of smaller businesses; they don’t really question what I want to spend within reason (and ultimately the limiting factor is that if it goes over a certain amount, it goes to the board of directors for approval).
Also also remember that Microsoft licensing is now per core, so you have to be careful to walk that line. I uh, have a lot of cores in my hosts, and they were licensed per socket when we started.
Hybrid 365 is so damn stupid, especially when 1 out of every 10 accounts you attempt to create ends up creating both a Cloud and On-Prem mailbox, even when you have never assigned an Exchange Online license before.
Hybrid 365 is a headache.
The best part is that they market it as an option for long-term migration but they don't tell you that you can't remove Exchange at the end and keep ADSync running. Genius move leaving key features on-prem only which requires you to keep the Exchange server, or run in an unsupported setup.
Because if you're going to attempt to squeeze that big black monster into your slot you will need to be able to take at least 12 inches or else you're going to have a bad time...
I will say that if the company is only willing to spend about $5k on an Active Directory implementation, I’d consider getting one physical box with good redundancy (good RAID, dual power supplies, etc) in its own hardware rather than trying to squeeze a bunch of VMs on there just for the sake of virtualization.
I’d also question an installer who wants to do that for $5k. In my opinion, to kind of go along with what Feral was saying, if you’re getting into the virtualization game, I think you should be at two hosts and a NAS to begin with if you want to do it right. I haven’t specced out low cost VMware hosts in a while but I think you’d be in the $20k range total, maybe less if you compromise in a couple areas. But I do have to admit that my virtualization budget is rather good compared to a lot of smaller businesses; they don’t really question what I want to spend within reason (and ultimately the limiting factor is that if it goes over a certain amount, it goes to the board of directors for approval).
Also also remember that Microsoft licensing is now per core, so you have to be careful to walk that line. I uh, have a lot of cores in my hosts, and they were licensed per socket when we started.
Even if a client is cheap and doesn't want to spend $$$ on a big virtualization setup, I would still much rather set up a $5K server and virtualize it on a host for the ease of moving it if I have to in the future. IMO, copying a VHD to a fresh host install is easier than doing a restore of a server. It also allows more freedom with support if the VM goes down as I can still remote into the host if it is configured with external DNS servers.
Because if you're going to attempt to squeeze that big black monster into your slot you will need to be able to take at least 12 inches or else you're going to have a bad time...
I'm working with maybe 200ish desktop/laptops and probably under a thousand total users (Most have iPads, but we have a few labs) so we're definitely not huge. Making sure this is as inexpensive as possible is going to be a high priority, especially since I have to convince people who will barely understand what it does that it's actually worth doing. Setting up one server with VMs that we can easily spread out the following year might be ideal for now, but I'll probably bring some "ideal" quotes and some "bare minimum" quotes and see what I can eek out of 'em.
I will say that if the company is only willing to spend about $5k on an Active Directory implementation, I’d consider getting one physical box with good redundancy (good RAID, dual power supplies, etc) in its own hardware rather than trying to squeeze a bunch of VMs on there just for the sake of virtualization.
I’d also question an installer who wants to do that for $5k. In my opinion, to kind of go along with what Feral was saying, if you’re getting into the virtualization game, I think you should be at two hosts and a NAS to begin with if you want to do it right. I haven’t specced out low cost VMware hosts in a while but I think you’d be in the $20k range total, maybe less if you compromise in a couple areas. But I do have to admit that my virtualization budget is rather good compared to a lot of smaller businesses; they don’t really question what I want to spend within reason (and ultimately the limiting factor is that if it goes over a certain amount, it goes to the board of directors for approval).
Also also remember that Microsoft licensing is now per core, so you have to be careful to walk that line. I uh, have a lot of cores in my hosts, and they were licensed per socket when we started.
Even if a client is cheap and doesn't want to spend $$$ on a big virtualization setup, I would still much rather set up a $5K server and virtualize it on a host for the ease of moving it if I have to in the future. IMO, copying a VHD to a fresh host install is easier than doing a restore of a server. It also allows more freedom with support if the VM goes down as I can still remote into the host if it is configured with external DNS servers.
Exactly. You don't need to spend $20k on a virtualization host. For a small shop, you can do it with $5k per host on hardware, plus nothing for the free VMware ESXi license, or do Hyper-V. You just have to make sure your OSes are properly licensed.
I wouldn't call 200 computers and 1000 users a small shop, though. To me, that's midsized.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
Pop quiz, sysadmin thread. This should be an easy one.
It's 7:00am. There are four IT staff in the office and there's a crisis brewing. Key points:
Multiple users are unable to log in to their computers.
Affected computers do not respond to ping.
A couple of your servers also seem to be down, not responding to ping.
You checked in with one of your branch offices and they're fine. Not all of your branch offices are open yet though. So for now you're not sure which locations are affected, but you know you have at least one branch office that is up and running.
When a computer is rebooted, it drops off the network and stops responding to ping.
What would be your first avenue of investigation?
Feral on
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
I have a pretty strong belief that if you can't afford two physical servers, you can't afford Active Directory.
I recognize we'll probably disagree on that.
This is also saying the network security that comes with GPO and user authentication has an entry level price tag of 10-20k, which I disagree with.
You can have AD on the cheap, you just need to understand that your business continuity plan is not geared for fast recovery. You’re going to have more single points of failure. That doesn’t mean you can’t afford gpo and network storage. Better a virtual server running on a single raid’ed host getting backed up to carbonite or rotating USB drives, than shit spread around on workgrouped PCs.
Pop quiz, sysadmin thread. This should be an easy one.
It's 7:00am. There are four IT staff in the office and there's a crisis brewing. Key points:
Multiple users are unable to log in to their computers.
Affected computers do not respond to ping.
A couple of your servers also seem to be down, not responding to ping.
You checked in with one of your branch offices and they're fine. Not all of your branch offices are open yet though. So for now you're not sure which locations are affected, but you know you have at least one branch office that is up and running.
When a computer is rebooted, it drops off the network and stops responding to ping.
What would be your first avenue of investigation?
DHCP
Edit: wait, servers too. DNS
Cog on
+4
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
DHCP was my thought as well.
"Stops responding", not "can't resolve"? DNS could be cached, I guess.
I'd check a workstation, see if it can get an IP, see if it can ping external addresses, etc.
Somebody plugged an apple airport into the network.
life's a game that you're bound to lose / like using a hammer to pound in screws
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
I will say that if the company is only willing to spend about $5k on an Active Directory implementation, I’d consider getting one physical box with good redundancy (good RAID, dual power supplies, etc) in its own hardware rather than trying to squeeze a bunch of VMs on there just for the sake of virtualization.
I’d also question an installer who wants to do that for $5k. In my opinion, to kind of go along with what Feral was saying, if you’re getting into the virtualization game, I think you should be at two hosts and a NAS to begin with if you want to do it right. I haven’t specced out low cost VMware hosts in a while but I think you’d be in the $20k range total, maybe less if you compromise in a couple areas. But I do have to admit that my virtualization budget is rather good compared to a lot of smaller businesses; they don’t really question what I want to spend within reason (and ultimately the limiting factor is that if it goes over a certain amount, it goes to the board of directors for approval).
Also also remember that Microsoft licensing is now per core, so you have to be careful to walk that line. I uh, have a lot of cores in my hosts, and they were licensed per socket when we started.
Even if a client is cheap and doesn't want to spend $$$ on a big virtualization setup, I would still much rather set up a $5K server and virtualize it on a host for the ease of moving it if I have to in the future. IMO, copying a VHD to a fresh host install is easier than doing a restore of a server. It also allows more freedom with support if the VM goes down as I can still remote into the host if it is configured with external DNS servers.
One of the cheaper redundant hardware virtualization products I've seen is Starwind HCA. Essentially a pair of Dell warranty-backed workstation-class towers with redundant power and hardware RAID, using a "Virtual SAN" software to synchronize the local data storage volumes between the two physical servers using direct-attached 10 Gb cables for the storage traffic so you don't need more expensive 10Gb network switches. Uses Hyper-V (or ESX, but for this case, he'd want to stay with the Hyper-V because the free ESX doesn't do migration/vmotion without vCenter). The VM's live on the synchronized volumes, so if you have a hardware failure, you just start them back up on the second box.
Now in my opinion, these are fine for small remote office applications where you are only planning to run a handful of VM's. Active Directory/DNS servers require very little compute and memory so if that's all this is it should be fine, but don't expect to be running a whole raft of VM's off of such a setup or real heavy application loads, you'd probably want better servers for that.
One bit of complication is that the Hyper-V clustering that you need to support migration and failover between the servers requires AD. Since there's no AD environment, you'd have to set up one server, get your AD VM's spun up on that, then set up the second server and clustering.
Note that you'd still need a backup solution. Synchronized data storage is not backup.
Just remember that half the people you meet are below average intelligence.
I have a pretty strong belief that if you can't afford two physical servers, you can't afford Active Directory.
I recognize we'll probably disagree on that.
This is also saying the network security that comes with GPO and user authentication has an entry level price tag of 10-20k, which I disagree with.
You can have AD on the cheap, you just need to understand that your business continuity plan is not geared for fast recovery. You’re going to have more single points of failure. That doesn’t mean you can’t afford gpo and network storage. Better a virtual server running on a single raid’ed host getting backed up to carbonite or rotating USB drives, than shit spread around on workgrouped PCs.
Nobody will understand this except you. When it eventually fails - and it will - nobody else is going to remember the very sober conversation where you clearly explained the situation to your CFO.
They'll just be cursing your name and spreading rumors about how bad their IT is.
They might even tell you to tear the server down permanently because it isn't worth the liability.
Meanwhile you've enabled a pattern of IT <-> MGMT interactions where they're able to get you to sacrifice critical values like "reliability" and "uptime" by low-balling you. You're allowing them to take out technical debt like a bartender giving shots to an alcoholic.
And your own stress and anxiety will be amplified on a day to day basis because you have this single server that you're afraid to touch because taking it down will fuck up your whole network.
You'll spend more time working late nights and early mornings because you can't do routine maintenance or invasive troubleshooting during business hours.
It's not worth it.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
There were four IT staff working this problem: one helpdesk, two admins, and one manager. They texted me early on to tell me what was up but I didn't see the texts because I was asleep.
By the time I saw the texts and jumped on, they'd be investigating it for over an hour, still didn't have any idea what the problem was. Their working hypothesis was "a bad Windows Update was installed on the affected PCs" and somebody opened a trouble ticket with Microsoft over it.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
Pop quiz, sysadmin thread. This should be an easy one.
It's 7:00am. There are four IT staff in the office and there's a crisis brewing. Key points:
Multiple users are unable to log in to their computers.
Affected computers do not respond to ping.
A couple of your servers also seem to be down, not responding to ping.
You checked in with one of your branch offices and they're fine. Not all of your branch offices are open yet though. So for now you're not sure which locations are affected, but you know you have at least one branch office that is up and running.
When a computer is rebooted, it drops off the network and stops responding to ping.
What would be your first avenue of investigation?
There's half a dozen unanswered questions to tell for sure, but given this thread, DNS. Computers can't locate the domain, switch their firewall profile to Public and start blocking ping.
Just remember that half the people you meet are below average intelligence.
I was thinking DHCP but I had no idea if it's common practice nowadays to run servers on DHCP or not so I kept my mouth shut and huddled in the corner.
I'd probably first check physical hard before going to DHCP just in case it's a legit failing switch, but then DHCP, then DNS, then active directory fuckmuppetry.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
I was thinking DHCP but I had no idea if it's common practice nowadays to run servers on DHCP or not so I kept my mouth shut and huddled in the corner.
Yes and no, you can do a reservation for the IP with mac address but I prefer to just set it as static IP if I know it's supposed to be static.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
I was thinking DHCP but I had no idea if it's common practice nowadays to run servers on DHCP or not so I kept my mouth shut and huddled in the corner.
Yes and no, you can do a reservation for the IP with mac address but I prefer to just set it as static IP if I know it's supposed to be static.
DHCP reservation has some minor management benefit but also adds another point of failure as demonstrated above.
Just remember that half the people you meet are below average intelligence.
Pop quiz, sysadmin thread. This should be an easy one.
It's 7:00am. There are four IT staff in the office and there's a crisis brewing. Key points:
Multiple users are unable to log in to their computers.
Affected computers do not respond to ping.
A couple of your servers also seem to be down, not responding to ping.
You checked in with one of your branch offices and they're fine. Not all of your branch offices are open yet though. So for now you're not sure which locations are affected, but you know you have at least one branch office that is up and running.
When a computer is rebooted, it drops off the network and stops responding to ping.
Yeah I've been burnt on physical hardware enough for shit like this where I'd absolutely triple check that the first time before going on the wild goose chase.
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
+1
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
There were four IT staff working this problem: one helpdesk, two admins, and one manager. They texted me early on to tell me what was up but I didn't see the texts because I was asleep.
By the time I saw the texts and jumped on, they'd be investigating it for over an hour, still didn't have any idea what the problem was. Their working hypothesis was "a bad Windows Update was installed on the affected PCs" and somebody opened a trouble ticket with Microsoft over it.
I'd ask if y'all were hiring, if not for your other posts!
There were four IT staff working this problem: one helpdesk, two admins, and one manager. They texted me early on to tell me what was up but I didn't see the texts because I was asleep.
By the time I saw the texts and jumped on, they'd be investigating it for over an hour, still didn't have any idea what the problem was. Their working hypothesis was "a bad Windows Update was installed on the affected PCs" and somebody opened a trouble ticket with Microsoft over it.
There were four IT staff working this problem: one helpdesk, two admins, and one manager. They texted me early on to tell me what was up but I didn't see the texts because I was asleep.
By the time I saw the texts and jumped on, they'd be investigating it for over an hour, still didn't have any idea what the problem was. Their working hypothesis was "a bad Windows Update was installed on the affected PCs" and somebody opened a trouble ticket with Microsoft over it.
Well what do I win for being right first?
A free game from Humble Bundle! Check your PMs.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
So this isn't hard to do but I need to do it a lot and it annoys me every time, EVERYONE else uses .pem for ssl microsoft, fuck you stop making me convert them to pfx and just use pem you asshats. IJUFOIHJFEUIHFGEIUOYGIUYADE. Leave pfx support for people obsessed with security but just let me import fucking pems.
Posts
I did a write up for somebody in a similar situation. Lemme see if I can find it.
$5k is not unreasonable. I'd prefer a slightly higher budget personally
the "no true scotch man" fallacy.
This person was running on-premises Exchange, which you aren't doing, but the principles still apply.
Use redundancy and virtualization. You will thank yourself later.
Two physical servers, each running a virtualization platform. (Either VMware ESXi free version or Microsoft Hyper-V work, and both are affordable.)
One Active Directory server on each physical box. (Each Active Directory server will also be a DNS server. It's usually advisable you use Microsoft DHCP as well.)
Your software cost will be a little less because you're not doing Exchange. But you will also need to think about a rack or cage, UPS and power distribution, possibly cooling.
the "no true scotch man" fallacy.
backupsssssss
Eh
Who needs em
:rotate:
the "no true scotch man" fallacy.
If money is a SERIOUS issue, one physical server is an option. Not the greatest option, but budget constraints are real.
I recognize we'll probably disagree on that.
the "no true scotch man" fallacy.
I’d also question an installer who wants to do that for $5k. In my opinion, to kind of go along with what Feral was saying, if you’re getting into the virtualization game, I think you should be at two hosts and a NAS to begin with if you want to do it right. I haven’t specced out low cost VMware hosts in a while but I think you’d be in the $20k range total, maybe less if you compromise in a couple areas. But I do have to admit that my virtualization budget is rather good compared to a lot of smaller businesses; they don’t really question what I want to spend within reason (and ultimately the limiting factor is that if it goes over a certain amount, it goes to the board of directors for approval).
Also also remember that Microsoft licensing is now per core, so you have to be careful to walk that line. I uh, have a lot of cores in my hosts, and they were licensed per socket when we started.
This is a clickable link to my Steam Profile.
The best part is that they market it as an option for long-term migration but they don't tell you that you can't remove Exchange at the end and keep ADSync running. Genius move leaving key features on-prem only which requires you to keep the Exchange server, or run in an unsupported setup.
Even if a client is cheap and doesn't want to spend $$$ on a big virtualization setup, I would still much rather set up a $5K server and virtualize it on a host for the ease of moving it if I have to in the future. IMO, copying a VHD to a fresh host install is easier than doing a restore of a server. It also allows more freedom with support if the VM goes down as I can still remote into the host if it is configured with external DNS servers.
XBL:Phenyhelm - 3DS:Phenyhelm
Exactly. You don't need to spend $20k on a virtualization host. For a small shop, you can do it with $5k per host on hardware, plus nothing for the free VMware ESXi license, or do Hyper-V. You just have to make sure your OSes are properly licensed.
I wouldn't call 200 computers and 1000 users a small shop, though. To me, that's midsized.
the "no true scotch man" fallacy.
It's 7:00am. There are four IT staff in the office and there's a crisis brewing. Key points:
Multiple users are unable to log in to their computers.
Affected computers do not respond to ping.
A couple of your servers also seem to be down, not responding to ping.
You checked in with one of your branch offices and they're fine. Not all of your branch offices are open yet though. So for now you're not sure which locations are affected, but you know you have at least one branch office that is up and running.
When a computer is rebooted, it drops off the network and stops responding to ping.
What would be your first avenue of investigation?
the "no true scotch man" fallacy.
This is also saying the network security that comes with GPO and user authentication has an entry level price tag of 10-20k, which I disagree with.
You can have AD on the cheap, you just need to understand that your business continuity plan is not geared for fast recovery. You’re going to have more single points of failure. That doesn’t mean you can’t afford gpo and network storage. Better a virtual server running on a single raid’ed host getting backed up to carbonite or rotating USB drives, than shit spread around on workgrouped PCs.
DHCP
Edit: wait, servers too. DNS
"Stops responding", not "can't resolve"? DNS could be cached, I guess.
I'd check a workstation, see if it can get an IP, see if it can ping external addresses, etc.
fuck up once and you break your thumb / if you're happy at all then you're god damn dumb
that's right we're on a fucked up cruise / God is dead but at least we have booze
bad things happen, no one knows why / the sun burns out and everyone dies
One of the cheaper redundant hardware virtualization products I've seen is Starwind HCA. Essentially a pair of Dell warranty-backed workstation-class towers with redundant power and hardware RAID, using a "Virtual SAN" software to synchronize the local data storage volumes between the two physical servers using direct-attached 10 Gb cables for the storage traffic so you don't need more expensive 10Gb network switches. Uses Hyper-V (or ESX, but for this case, he'd want to stay with the Hyper-V because the free ESX doesn't do migration/vmotion without vCenter). The VM's live on the synchronized volumes, so if you have a hardware failure, you just start them back up on the second box.
Now in my opinion, these are fine for small remote office applications where you are only planning to run a handful of VM's. Active Directory/DNS servers require very little compute and memory so if that's all this is it should be fine, but don't expect to be running a whole raft of VM's off of such a setup or real heavy application loads, you'd probably want better servers for that.
One bit of complication is that the Hyper-V clustering that you need to support migration and failover between the servers requires AD. Since there's no AD environment, you'd have to set up one server, get your AD VM's spun up on that, then set up the second server and clustering.
Note that you'd still need a backup solution. Synchronized data storage is not backup.
Nobody will understand this except you. When it eventually fails - and it will - nobody else is going to remember the very sober conversation where you clearly explained the situation to your CFO.
They'll just be cursing your name and spreading rumors about how bad their IT is.
They might even tell you to tear the server down permanently because it isn't worth the liability.
Meanwhile you've enabled a pattern of IT <-> MGMT interactions where they're able to get you to sacrifice critical values like "reliability" and "uptime" by low-balling you. You're allowing them to take out technical debt like a bartender giving shots to an alcoholic.
And your own stress and anxiety will be amplified on a day to day basis because you have this single server that you're afraid to touch because taking it down will fuck up your whole network.
You'll spend more time working late nights and early mornings because you can't do routine maintenance or invasive troubleshooting during business hours.
It's not worth it.
the "no true scotch man" fallacy.
Also make your argument and record their consent in email form so later you can print it out and slap them with it.
There were four IT staff working this problem: one helpdesk, two admins, and one manager. They texted me early on to tell me what was up but I didn't see the texts because I was asleep.
By the time I saw the texts and jumped on, they'd be investigating it for over an hour, still didn't have any idea what the problem was. Their working hypothesis was "a bad Windows Update was installed on the affected PCs" and somebody opened a trouble ticket with Microsoft over it.
the "no true scotch man" fallacy.
There's half a dozen unanswered questions to tell for sure, but given this thread, DNS. Computers can't locate the domain, switch their firewall profile to Public and start blocking ping.
This makes much more sense from a contractor/MSP point of view.
It's career suicide from an internal IT point of view.
Yes and no, you can do a reservation for the IP with mac address but I prefer to just set it as static IP if I know it's supposed to be static.
DHCP reservation has some minor management benefit but also adds another point of failure as demonstrated above.
ARP table is fubared. Reboot switches.
Some of our dev servers are DHCP.
But of course there's the occasional server running production workloads in dev that never got properly migrated over.
DHCP doesn't have to be a single point of failure. I've been meaning to set up redundant DHCP but, yknow, malcolminthemiddlelightbulb.avi
the "no true scotch man" fallacy.
I'd ask if y'all were hiring, if not for your other posts!
Well what do I win for being right first?
A free game from Humble Bundle! Check your PMs.
the "no true scotch man" fallacy.
if it's not Deer Hunter 2019 I give up on this thread
It’s not.
It’s better.
Well played, @Feral