I mean I wouldn't pay for it, but it'd be cool actually having win95 usb edition on CD someday when I want to take a trip down memory lane without having to find it on the tubes.
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
Yo dawg, we found DOS 6.22 floppies the other day when cleaning out our software cabinet. We kept them.
jungleroomxIt's never too many graves, it's always not enough shovelsRegistered Userregular
edited November 2017
Welp, we finally got our pristine virtual server environment enabled, denied access to developers, and started on setting up a default customer setup for reference.
About 10 minutes in, we get an email from a high muckity-muck that we are no longer allowed to do it for unspecified reasons.
Oh, look at that, an it business analyst spot that pays double what I do now and I fit the quals pretty well.
*updates resume*
jungleroomx on
+8
Options
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
Welp, we finally got our pristine virtual server environment enabled, denied access to developers, and started on setting up a default customer setup for reference.
About 10 minutes in, we get an email from a high muckity-muck that we are no longer allowed to do it for unspecified reasons.
Welp, we finally got our pristine virtual server environment enabled, denied access to developers, and started on setting up a default customer setup for reference.
About 10 minutes in, we get an email from a high muckity-muck that we are no longer allowed to do it for unspecified reasons.
They have 24 TB of storage in DAS off their ESX hosts but only 12TB of it was actually provisioned. Failed drive in the array. No hot spares were configured.
Everything was out of warranty
still backing up to a single-tape drive
One of their DCs is in a site that has an assigned subnet but it has an IP that doesn't match that subnet, so replication is all fucked up
There are random printers and folders shared of off random file servers and dcs everywhere. Users complain that network resources and mapped drives randomly appear and disappear
They have a DHCP scope that has a fucking absurd amount of reservations in it including apparently just some random workstations, reservations made on IP addresses outside the goddamn scope, and a reservation for one of their file servers
I logged in to said file server and checked... yep, sure enough, it's set to DHCP
but what's this? A second NIC? Set to a static IP address
within the fucking DHCP scope
that has no reservation or exclusion :rotate:
lets just see... Yeah, Bad_address. It's conflicting with something. No wonder your shares and printers randomly disappear.
It might just be easier to advise them to burn the building down.
The previous IT guy had handed out instructions for mapping drives after connecting to the VPN, but one of the drives would never map for anyone and he never figured it out. I saw in 5 seconds that he mispelled the name of the shared folder on the instructions.
The previous IT guy had handed out instructions for mapping drives after connecting to the VPN, but one of the drives would never map for anyone and he never figured it out. I saw in 5 seconds that he mispelled the name of the shared folder on the instructions.
:tell_me_more:
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
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
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
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
Was this on 10/31? I think that's the end of Dell's 3rd Quarter, so it may be related to sales quotas or discounts that may have been expiring.
Just remember that half the people you meet are below average intelligence.
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
Was this on 10/31? I think that's the end of Dell's 3rd Quarter, so it may be related to sales quotas or discounts that may have been expiring.
11/1
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 replaced one of our old Catalyst 4500s with a new 3850 stack last night. Specifically I set both to HSRP with "standby ip" then forced the 3850 stack to HSRP active.
Remember from my prior adventures that the spanning tree configs on those old 4500s were completely fucked. BTW, so were the EIGRP configs. EIGRP was completely broken because some genius basically flipped a coin to decide which routers across the old organization should be stubs, which shouldn't, which routers had redistribute static, which shouldn't, etc.
The 4500s also had VLANs and subnets configured that were no longer in use, and a ton of other legacy garbage.
So before the cutover last night, I migrated over only the VLANs and subnets and static routes we still use. I dropped EIGRP and used static for everything. None of this was well documented so it took a lot of educated guesswork and wiresharking over the last several months to figure out which subnets/routes/etc were safe to abandon.
After changing the active router to the 3850 stack, I pinged and tracerouted everything I could possibly think of. Tracerouted workstation to server, server to workstation, workstation to Internet, remote office to datacenter, server to VOIP phones, etc. I played with packet sizes, got my laptop on a mobile hotspot and logged in with VPN, did some large downloads from the Internet, etc.
Everything looked fantastic.
So I unplugged old 4500 and powered it down. Re-ran a few ping and tracerouted tests and everything still looked good, so I texted my boss and a couple of coworkers and then went home.
...until this morning... (continued...)
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.
This morning I got the following text message from a coworker:
"PCs on the 169 subnet can't get on network"
We don't have a 169 subnet.
Now if you know the first fucking thing about Windows you know that if it can't get a DHCP address, it will autoassign an address starting with 169. Why literally nobody else in the department at the time couldn't identify 169.xxx.xxx.xxx as a Windows autoconfiguration address, and therefore a DHCP problem, is a story in and of itself. But this is a story about Cisco, not my coworkers, so I digress.
I figured it was something stupid, like I forgot to declare ip helper-address on a VLAN or something like that. So I remoted in from home and looked at the 3850 configs...
...nope, I remembered all my ip helper-addresses.
So I did the usual debug commands like debug ip dhcp server packets and debug ip udp, and I also logged into our DHCP server to see if it had any interesting events.
debug ip dhcp server packets produced no output. At all. Zilch zero nada.
debug ip udp showed DHCP request packets (UDP port 67) coming in from workstations, but none of them were getting forwarded.
DHCP server logs (predictably, at this point) showed no DHCP packets being received.
So I drove into the office, opened a case with Cisco support, and continued to troubleshoot. We're now about an hour into the problem with multiple people across multiple departments unable to get on the network, and nobody else in my department knows enough about TCP/IP in Windows to identify a 169. IP address but i'm digressing again sorry...
While waiting for Cisco to call me back, I comb through the old 4500 configs to see if there's anything I missed. Any DHCP-related commands or any routing-related commands. Nope, nothing. I also looked through some of our other Catalysts across the network to see if there's anything configured on those that I might have missed. Nope, not a thing.
More Googling and I come across a forum post from somebody who had the same problem and said that the command service dhcp fixed it. So, fine, fuck it, I try it.
Suddenly the dhcp debug starts to display forwarding events and the DHCP server starts to receive DHCP requests. I can see workstations lighting up across the building.
Note that service dhcp is not declared on any other Catalyst in our network nor was it declared on the old 4500.
A little while later, Cisco calls me back and I run it by the tech and he's like
"You know, you're the second customer this week with that exact problem. You're right, you shouldn't have to do "service dhcp." It should be on by default. I think it's a bug in our 3850 firmware but I'm not sure."
...
cisco
cisco plz
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.
Why the hell is this one user profile acting like a roaming profile on this specific server even though it's not set as one and doesn't act this way on any other machine and no other profile acts this way on this specific server.
because windows
have you deleted/renamed the profile folder and deleted the associated registry keys* so it re-makes it from scratch?
Why the hell is this one user profile acting like a roaming profile on this specific server even though it's not set as one and doesn't act this way on any other machine and no other profile acts this way on this specific server.
because windows
have you deleted/renamed the profile folder and deleted the associated registry keys* so it re-makes it from scratch?
The thing is, none of those on this specific machine exist unless I'm explicitly signed in as the user account. I sign out from the account and log in as an administrator account and they disappear from the registry. It's not set to act this way and doesn't act this way anywhere else. And, again, it's the only profile affected by this problem on this machine.
Did someone mistakenly stick their account in the local Guests (or Domain Guests) group? Members in those groups will have their profile folders and registry keys deleted on logout.
In case you were wondering, we found the cause. Remote Desktop Services has a "feature" you can turn on for collections called User Profile Disks. This setting had been turned on back in April by someone and they were starting to corrupt, which was starting to spread to other profiles. Turning this setting off fixed it.
It's a feature in that user profile disks are used if you have multiple RDS hosts in a farm and want your users to be able to log into any of the hosts while keeping their same profile.
Just remember that half the people you meet are below average intelligence.
The previous IT guy had handed out instructions for mapping drives after connecting to the VPN, but one of the drives would never map for anyone and he never figured it out. I saw in 5 seconds that he mispelled the name of the shared folder on the instructions.
:tell_me_more:
This place wants the work to fix the major issues to be done on a weekend so they don't have business hour downtime.
Would be a pretty power move to just flip them to a whole new domain over the weekend and have them come in and everything is fucking flawless.
Would be more likely they come in and i'm passed out on the server room floor and the esx host is purple screened and nothing works and there's a thousand photo copies of my ass in the recycling and the trash can in the men's room is on fire
+4
Options
RandomHajileNot actually a SnatcherThe New KremlinRegistered Userregular
Welp, we finally got our pristine virtual server environment enabled, denied access to developers, and started on setting up a default customer setup for reference.
About 10 minutes in, we get an email from a high muckity-muck that we are no longer allowed to do it for unspecified reasons.
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
Was this on 10/31? I think that's the end of Dell's 3rd Quarter, so it may be related to sales quotas or discounts that may have been expiring.
11/1
They had a promotion for extending warranty on old servers and it was through 11/1 according to our warranty specialist dude, yeah.
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
Was this on 10/31? I think that's the end of Dell's 3rd Quarter, so it may be related to sales quotas or discounts that may have been expiring.
11/1
They had a promotion for extending warranty on old servers and it was through 11/1 according to our warranty specialist dude, yeah.
New server though!
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
The previous IT guy had handed out instructions for mapping drives after connecting to the VPN, but one of the drives would never map for anyone and he never figured it out. I saw in 5 seconds that he mispelled the name of the shared folder on the instructions.
:tell_me_more:
This place wants the work to fix the major issues to be done on a weekend so they don't have business hour downtime.
That's literally not going to happen
It is literally impossible
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.
Yeah I always tell my boss, "sure we can limp along for several weeks with me fixing shit here and there, or I can rip the band aid off for 30-60 minutes during lunch"
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 mean, I can probably unfuck AD replication and Re-IP the servers that are in the DHCP scope, and get drive & printer mapping via GPO over a weekend.
The one that's really fucked though is the DC in the site in Sites & Services that has the IP that doesn't match the defined subnet. That's a bitch.
And it's not going to "fix all the problems" regardless. It might just eliminate some of the big pain points.
Nothing's going to "fix" things until we set up a whole new domain and jettison this old dumpster fire.
EDIT: Oh, and having everyone stop being local admins of everything. I can probably work that out, but that will cause some downtime because there's just no way to test everything.
Most of the time when we on-board a client into one of our managed infrastructure services, one of our pre-conditions is that we will be shit-canning your old AD. It's nearly never worth the effort, and almost everyone's domain sucks ass anyway.
Most of the time when we on-board a client into one of our managed infrastructure services, one of our pre-conditions is that we will be shit-canning your old AD. It's nearly never worth the effort, and almost everyone's domain sucks ass anyway.
I can't imagine doing that for every client but okay
If somebody tried to do that here they'd quickly end up in thisisfine.jpg
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.
Most of the time when we on-board a client into one of our managed infrastructure services, one of our pre-conditions is that we will be shit-canning your old AD. It's nearly never worth the effort, and almost everyone's domain sucks ass anyway.
I can't imagine doing that for every client but okay
If somebody tried to do that here they'd quickly end up in thisisfine.jpg
Statistically speaking, your AD environment is probably garbage.
Also we spin out template VMs and as soon as they boot up we have a enormous powershell script that installs AD DS and promotes the server and generates the OU structure and all of the general service accounts and things we always use. It goes from zero to fully functioning domain in about 20 minutes.
Most of the time when we on-board a client into one of our managed infrastructure services, one of our pre-conditions is that we will be shit-canning your old AD. It's nearly never worth the effort, and almost everyone's domain sucks ass anyway.
I can't imagine doing that for every client but okay
If somebody tried to do that here they'd quickly end up in thisisfine.jpg
Statistically speaking, your AD environment is probably garbage.
It was definitely garbage when I started.
At this point it's only about 20% garbage.
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.
You people and your Windows 95's. This is actually one of my more prized possessions among my IT/comptuer stuff. This was handed down to me by a friend of my dad years ago, as DOS 2.10 is a bit before my time. I first learned computers on DOS 6 and Win 3.1.
I put it in an imgur gallery since I took 12 pictures of it. It's the retail package of IBM Disk Operating System 2.10, provided by Microsoft. It includes a manual with all of the DOS commands.
Posts
"Click to enable Flash Player"
oh, you cads.
Impossible.
Anyone need a copy of Windows 95?
This is a clickable link to my Steam Profile.
About 10 minutes in, we get an email from a high muckity-muck that we are no longer allowed to do it for unspecified reasons.
Oh, look at that, an it business analyst spot that pays double what I do now and I fit the quals pretty well.
*updates resume*
This is a clickable link to my Steam Profile.
Ah ok, I can fix that.
this fucking place
They have 24 TB of storage in DAS off their ESX hosts but only 12TB of it was actually provisioned. Failed drive in the array. No hot spares were configured.
Everything was out of warranty
still backing up to a single-tape drive
One of their DCs is in a site that has an assigned subnet but it has an IP that doesn't match that subnet, so replication is all fucked up
There are random printers and folders shared of off random file servers and dcs everywhere. Users complain that network resources and mapped drives randomly appear and disappear
They have a DHCP scope that has a fucking absurd amount of reservations in it including apparently just some random workstations, reservations made on IP addresses outside the goddamn scope, and a reservation for one of their file servers
I logged in to said file server and checked... yep, sure enough, it's set to DHCP
but what's this? A second NIC? Set to a static IP address
within the fucking DHCP scope
that has no reservation or exclusion :rotate:
lets just see... Yeah, Bad_address. It's conflicting with something. No wonder your shares and printers randomly disappear.
It might just be easier to advise them to burn the building down.
the "no true scotch man" fallacy.
:tell_me_more:
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
Ok, now put your RDP window on your right screen so you can enjoy the annoyance you've just created for yourself.
Nah, I do all of my administration through Powershell sessions and RSAT.
People who only use GUIs should be punished.
the "no true scotch man" fallacy.
the "no true scotch man" fallacy.
dell guy called like
3 hours before it shipped
but it was like end of the day for us and I thought he was trying to upsell me on storage or some shit so I let it roll to voice mail
nope he wanted to upgrade our shipping and support for free but I needed to respond within like that 40 minute window
boss was upset but like bro I get a dozen sales calls a day I can't spend all day dealing with vendors, and a 40 minute window was not enough
We can wait an extra 4 days for the server, and it still baffles my mind why this guy can't upgrade our support component.. I can obviously add it after the fact.
Was this on 10/31? I think that's the end of Dell's 3rd Quarter, so it may be related to sales quotas or discounts that may have been expiring.
11/1
I replaced one of our old Catalyst 4500s with a new 3850 stack last night. Specifically I set both to HSRP with "standby ip" then forced the 3850 stack to HSRP active.
Remember from my prior adventures that the spanning tree configs on those old 4500s were completely fucked. BTW, so were the EIGRP configs. EIGRP was completely broken because some genius basically flipped a coin to decide which routers across the old organization should be stubs, which shouldn't, which routers had redistribute static, which shouldn't, etc.
The 4500s also had VLANs and subnets configured that were no longer in use, and a ton of other legacy garbage.
So before the cutover last night, I migrated over only the VLANs and subnets and static routes we still use. I dropped EIGRP and used static for everything. None of this was well documented so it took a lot of educated guesswork and wiresharking over the last several months to figure out which subnets/routes/etc were safe to abandon.
After changing the active router to the 3850 stack, I pinged and tracerouted everything I could possibly think of. Tracerouted workstation to server, server to workstation, workstation to Internet, remote office to datacenter, server to VOIP phones, etc. I played with packet sizes, got my laptop on a mobile hotspot and logged in with VPN, did some large downloads from the Internet, etc.
Everything looked fantastic.
So I unplugged old 4500 and powered it down. Re-ran a few ping and tracerouted tests and everything still looked good, so I texted my boss and a couple of coworkers and then went home.
...until this morning... (continued...)
the "no true scotch man" fallacy.
"PCs on the 169 subnet can't get on network"
We don't have a 169 subnet.
Now if you know the first fucking thing about Windows you know that if it can't get a DHCP address, it will autoassign an address starting with 169. Why literally nobody else in the department at the time couldn't identify 169.xxx.xxx.xxx as a Windows autoconfiguration address, and therefore a DHCP problem, is a story in and of itself. But this is a story about Cisco, not my coworkers, so I digress.
I figured it was something stupid, like I forgot to declare ip helper-address on a VLAN or something like that. So I remoted in from home and looked at the 3850 configs...
...nope, I remembered all my ip helper-addresses.
So I did the usual debug commands like debug ip dhcp server packets and debug ip udp, and I also logged into our DHCP server to see if it had any interesting events.
debug ip dhcp server packets produced no output. At all. Zilch zero nada.
debug ip udp showed DHCP request packets (UDP port 67) coming in from workstations, but none of them were getting forwarded.
DHCP server logs (predictably, at this point) showed no DHCP packets being received.
So I drove into the office, opened a case with Cisco support, and continued to troubleshoot. We're now about an hour into the problem with multiple people across multiple departments unable to get on the network, and nobody else in my department knows enough about TCP/IP in Windows to identify a 169. IP address but i'm digressing again sorry...
While waiting for Cisco to call me back, I comb through the old 4500 configs to see if there's anything I missed. Any DHCP-related commands or any routing-related commands. Nope, nothing. I also looked through some of our other Catalysts across the network to see if there's anything configured on those that I might have missed. Nope, not a thing.
More Googling and I come across a forum post from somebody who had the same problem and said that the command service dhcp fixed it. So, fine, fuck it, I try it.
Suddenly the dhcp debug starts to display forwarding events and the DHCP server starts to receive DHCP requests. I can see workstations lighting up across the building.
Note that service dhcp is not declared on any other Catalyst in our network nor was it declared on the old 4500.
A little while later, Cisco calls me back and I run it by the tech and he's like
"You know, you're the second customer this week with that exact problem. You're right, you shouldn't have to do "service dhcp." It should be on by default. I think it's a bug in our 3850 firmware but I'm not sure."
...
cisco
cisco plz
the "no true scotch man" fallacy.
In case you were wondering, we found the cause. Remote Desktop Services has a "feature" you can turn on for collections called User Profile Disks. This setting had been turned on back in April by someone and they were starting to corrupt, which was starting to spread to other profiles. Turning this setting off fixed it.
This place wants the work to fix the major issues to be done on a weekend so they don't have business hour downtime.
Would be a pretty power move to just flip them to a whole new domain over the weekend and have them come in and everything is fucking flawless.
This is a clickable link to my Steam Profile.
This is a clickable link to my Steam Profile.
New server though!
That's literally not going to happen
It is literally impossible
the "no true scotch man" fallacy.
They are completely insane if they think their problems can be fixed without business impact.
the "no true scotch man" fallacy.
The one that's really fucked though is the DC in the site in Sites & Services that has the IP that doesn't match the defined subnet. That's a bitch.
And it's not going to "fix all the problems" regardless. It might just eliminate some of the big pain points.
Nothing's going to "fix" things until we set up a whole new domain and jettison this old dumpster fire.
EDIT: Oh, and having everyone stop being local admins of everything. I can probably work that out, but that will cause some downtime because there's just no way to test everything.
I can't imagine doing that for every client but okay
If somebody tried to do that here they'd quickly end up in thisisfine.jpg
the "no true scotch man" fallacy.
Statistically speaking, your AD environment is probably garbage.
It was definitely garbage when I started.
At this point it's only about 20% garbage.
the "no true scotch man" fallacy.
I put it in an imgur gallery since I took 12 pictures of it. It's the retail package of IBM Disk Operating System 2.10, provided by Microsoft. It includes a manual with all of the DOS commands.
https://imgur.com/gallery/6OnlJ
XBL:Phenyhelm - 3DS:Phenyhelm