Q: AD Windows 2003 -- How to Reinstall in small office

[Sky]

Hi there: couldn't find a newsgroup for Win2003 -- but hope someone has the
answer to this one anyway as I think this probably affects both servers...


We had a server hard-drive fail last week, and the backup failed to generate
a catalog, so couldn't extract anything. Day in hell. But that's beside the
point...

Went out and bought a new set of drives, installed, formatted, and
reinstalled Win2003.

Now, you've got to understand that I am one of the persons that has no good
grasp on AD, no matter how many articles I read. They all seem to be geared
at some huge company, multiple trees, etc..., whereas we are only 10
stations grouped around one file/web server. All we really want is some
permissioning, and DNS/DHCP, VPN.

That said, installed AD, followed the prompts, and gave it the same tree
name as before, something like "mycompany.fr", instead of "mycompany.local",
in case that's relevant.
Anyway -- all looked much healthier on server when finished. Put the users
and their passwords back in, made some groups, then went to log the clients
back in...
Each was able to be correctly signed in, and the server was willing to let
them access files on the server -- but barely. From what I figured out in
the cryptic (WHY ARE THEY SO CRYPTIC!?!) log messages, it was saying "yeah,
you clients are members of the same domain, but you belong to a server
that's not me, so you're distant cousins...Distant cousins don't get as many
rights to use printers, etc. Plus I make some Excel and Outlook
installations crash for the fun of it when you try to run them...".

We finally figured out a way to get around this, by going to each station,
saving their Favorites, My Documents, Outlook folders, and then signing in
as local Admin to delete all the profiles and then create a new domain
profile from scratch.

Question 1:
Ok. Ten client stations is not that many, but still, this sounds like it
would be a nightmare in a larger office, so there has to be a simpler way of
reinstalling a new server if a total break down happens.

Question 2:
I assume that there is a way of backing up whatever AD is, and being able to
move it to new hardware if one wants to. But how????

Question 3:
That was last week. Now we noticed tonight that we've obviously lost somehow
a crucial file (HOW?!?), because when we try to edit Local DomainController
Settings, it comes up void/blank/error. can't find the ...forgot the
filename. Sorry. But from posts it appears that if I turn on DFS, it might
go away ...atleast try that first -- but when i go to take a look at DFS, it
sure doesn't look like that is what I want, as we are only ONE server...who
would i synchronize with??

So we are going to reinstall the whole server again hence back to questions
1 and 2... The only good news (sortof) is that i'm getting faster and faster
at this...First time, took me 4 days, now it takes me about 3 hours to get
it formatted, os installed, mail server installed, rights, DNS, DCHP. etc. I
still have no idea what the heck I am doing, but it generally works.......
But it's pathetic! I wish I UNDERSTOOD... Plus I can't believe that I am
spending this time keeping a server running. It's become a full time
job...Big Rant.

Any help and clear instructions on how to do the above 3 things would be
greatly appreciated... Losing my mind.
Sky

 

[Scott Harding - MS MVP]

You need a consultant. Many of these issues you are running into are totally
avoidable. Like having only 1 Domain Controller is a very bad idea. If you
had 2, even one that was a very basic machine you would not have to rebuild
your domain. Everytime you reinstall the Domain you create a completely new
domain and thus all your clients have to rejoin your domain and then they
lose their profiles and you start from scratch again. DFS has nothing to do
with your issues. Most of your issues are most likely related to
misconfigured DNS settings. Your Domain Controller needs to point to itself
for DNS and all your clients need to point to it in their DNS settings. Then
on your Internal DNS server you use forwarders to your ISP's DNS server and
your server caches lookups for your clients. DNS must be configured
correctly for Active Directory to work. And when you say AD, "whatever that
is", it certainly is funny but in reality you should not be in charge of
this if you do not know what you doing. Sorry and no offense intended. This
all could have been avoided with a little knowledge. What I would suggest is
to hire someone locally to get your systems up and running and then they can
give you some guidelines to help you manage the day to day operations. They
can also tell/show you how to make proper backups and make the drives on the
server Fault Tolerant. A single hard drive failure should really not bring
your whole network down. Most servers can keep running when a drive fails IF
you setup the proper Fault tolerance on the disks. You certainly should not
have to do this and really isn't the right place for use to train you on
every single aspect of your network. Also there are many ways to make a
server almost completely fault tolerant so downtime is minimized. If you
have specific errors we always will help but making up errors in your post
really does absolutely nothing for us to help you.

 

[Sky]

Dear Scott:

No problem: your assessment is probably correct. I personally would have
called it as it is: "Monkey at the keyboard"...

But in the real world......
We went the route of a consultant, had this guy come in and do it for us.
Before we knew it, we had spent about euros1000 on the actual product and
about 4000 on the consultant, 1000 on add on products. I mean the thing just
kept working for about 10 days, then we had to call him in, and the hours
kept piling up. Maybe it was bad luck, maybe he was bad...who knows.

But money is money. So the boss threw the guy out, turned around, and said,
you learn it -- this is too expensive, as a single machine it's eaten up the
IT budget for the whole year for the other 10 clients, and we cant afford to
buy anything else right now...

So that's what I am trying to do. Help out.


What we had thought, to avoid the total loss of everything next time is to
Ghost the C:\drive by the end of the weekend with the OS, the DNS, etc.
settings just as we want it, and then if anything fails again, swap the
drives and reboot -- should be exactly the same position, no?

So when you say "don't waste your time", I'm real sorry -- seriously. Not
your problem. Got that.

But we still have a problem, and understanding AD a bit better would be
great.


Where we stand, and if you are ready to try to close the vaste gaps:

DNS
I think I've figured out the basics of DHCP and DNS. The server remains at
192.168.0.2, it uses itself as preferred DNS, and then falls back on out web
providers. The gateway is set to 192.168.0.1.

DHCP
As for DHCP that too appears to work fine. I've dropped the initial range to
192.168.0.100 to 50. Certain sets of 50 are used for direct IP of computers,
printers, etc. Just took me a while to find out that we had to create a
service account that had rights to the DNS service... The part that never
seemed to gel well was the Optional Services: the part where you tell the
Server to broadcast to the clients extra info at login time. We've set it to
broadcast DNS, and very basics -- but there are so many options that look
very useful. For example, time server info, gateway, etc. We did try the
time server options -- but that just caused a plethora of log messages
saying that we're set up as time server, but not getting information from
higher up in the chain of servers(?)

FileServing, IIS, etc. -- those are all no problem as they are basic rights.

AD
Ok. This is where I run into walls.
As I said, we're a small company, able to afford one server this year. I
understand the basics of users and groups and rights. What I don't
understand is the business of topologies, trees, replication. and the way AD
ties in to DNS...

With only one server, we're not a forest. We're a plank at best...so the
books we've bought from MS Press are wildly not the right scope.

What I would like to know is:
a) It's a single db file/dir? Where is it kept? Can it be backed up? Can it
be restored? This is a very serious matter: we can't be at the mercy of a
rights system that we have no rights to.... if the files we own are being
controlled by AD, we need to know how to manage this database. (If the
guards are guarding Rome, who is guarding the guards...etc.)
b) As we stated in the last post, is there no way to merge two domains? Or
merge the clients of an old one into a new one without blowing out their
profiles? None what so ever?
c) In a small company, 10 stations, what should the tree be called? The
Wizard offers you choice a) YourCompany.local, or b) one could do
"yourcompany.com" ... The world wide DNS services are pointing
"yourcompany.com" to the router here -- but we're thinking of hosting the
website somewhere else later.
d) If you build the AD tree by hand, you can call it what you want -- but if
you let the wizard do it, it gets called something like "Your First
Site"...can that be renamed? Is this visible name important? Probably yes,
because it's somehow tied into the DNS at some point, no?

Replication
That part doesn't make sense to me at all. We certainly are not in a
position to purchase 2 servers -- if I understand your suggestion. Are you
telling me that there are no small offices running everything off of one
server? That that is considered bad practice? I'm talking about the
thousands of SOHO /small offices across the country. Are you really telling
me that every home network with rights has to have two servers or they have
to re-install their client profiles everytime they buy a new server to
upgrade? Serious?!?!






PS:
As for the monkey thing...I agree -- I'm slow, and don't look to bright
under current circumstances in a forum of obviously bright MVP folks -- but
it was a monkey who was first to get into space, with a little help.
Care to help?

 

[Enkidu]

Hi Sky,

Scott is correct, it would be better to have someone who knows what
they are doing do it, but if that's not possible, read on. I've put my
comments in line.

Cheers,

Cliff

(MVP)

Dear Scott:

No problem: your assessment is probably correct. I personally would have
called it as it is: "Monkey at the keyboard"...

But in the real world......
We went the route of a consultant, had this guy come in and do it for us.
Before we knew it, we had spent about euros1000 on the actual product and
about 4000 on the consultant, 1000 on add on products. I mean the thing just
kept working for about 10 days, then we had to call him in, and the hours
kept piling up. Maybe it was bad luck, maybe he was bad...who knows.

Mmm. sounds a bit heavy, but who knows?

But money is money. So the boss threw the guy out, turned around, and said,
you learn it -- this is too expensive, as a single machine it's eaten up the
IT budget for the whole year for the other 10 clients, and we cant afford to
buy anything else right now...

So that's what I am trying to do. Help out.

Good luck! I don't know if you *like* this sort of thing, but if you
do, it's scary and it can be fun. The wins are great. The losses are
horrible. But I wouldn't be in any other business.

What we had thought, to avoid the total loss of everything next time is to
Ghost the C:\drive by the end of the weekend with the OS, the DNS, etc.
settings just as we want it, and then if anything fails again, swap the
drives and reboot -- should be exactly the same position, no?

I would not ghost a DC, especially one that is the *only* one. I would
ensure that it is backed up and and I would check the backup often
(restore a directory to an alternate location). I'd ensure that I
backed up the System State.

You really need a second DC like Scott says. I'd buy a cheap medium
spec box as a second DC if possible - I really believe that it would
save your life if the primary machine should fail again.

So when you say "don't waste your time", I'm real sorry -- seriously. Not
your problem. Got that.

Scott was giving good advice. It may not be appropriate for you, but
it was good advice. There are resources you can tap - friends and
relations in the business, local MVPs with maybe a little spare time.
Local newsgroups.

But we still have a problem, and understanding AD a bit better would be
great.

Where we stand, and if you are ready to try to close the vaste gaps:

DNS
I think I've figured out the basics of DHCP and DNS. The server remains at
192.168.0.2, it uses itself as preferred DNS, and then falls back on out web
providers. The gateway is set to 192.168.0.1.
Sounds good.

DHCP
As for DHCP that too appears to work fine. I've dropped the initial range to
192.168.0.100 to 50. Certain sets of 50 are used for direct IP of computers,
printers, etc. Just took me a while to find out that we had to create a
service account that had rights to the DNS service... The part that never
seemed to gel well was the Optional Services: the part where you tell the
Server to broadcast to the clients extra info at login time. We've set it to
broadcast DNS, and very basics -- but there are so many options that look
very useful. For example, time server info, gateway, etc. We did try the
time server options -- but that just caused a plethora of log messages
saying that we're set up as time server, but not getting information from
higher up in the chain of servers(?)

Be aware that not all Options are supported by clients. DHCP can
certainly deliver the basic necessities - IP address, DNS IPs, and
gateway IPs. I wouldn't worry about time services - by default XP and
2000 clients use the DC as time source (the one with the PDC Emulator
role if more than 1 DC).

FileServing, IIS, etc. -- those are all no problem as they are basic rights.

AD
Ok. This is where I run into walls.
As I said, we're a small company, able to afford one server this year. I
understand the basics of users and groups and rights. What I don't
understand is the business of topologies, trees, replication. and the way AD
ties in to DNS...

With only one server, we're not a forest. We're a plank at best...so the
books we've bought from MS Press are wildly not the right scope.

If you have one DC, you are the root server in the root domain of a
one domain tree in a single tree forest! Topology, replications,
trusts are all irrelevant to you at this time.

What I would like to know is:
a) It's a single db file/dir? Where is it kept? Can it be backed up? Can it
be restored? This is a very serious matter: we can't be at the mercy of a
rights system that we have no rights to.... if the files we own are being
controlled by AD, we need to know how to manage this database. (If the
guards are guarding Rome, who is guarding the guards...etc.)

Whoa! Don't panic! It's mostly a single file, but don't worry about
that, since you will *never* need to touch it. There are AD tools for
everything. You will find a tool on your system called NTBackup. Use
this to back up *everything* preferably on a daily (or nightly!)
basis. It will back all the drives and a thing called "System State".
On a DC that includes AD.

Users are authenticated though AD. Files and directories are protected
by permissions. Permissions are granted or denied to users.
Permissions information is kept in the file system not AD. And when
you restore from a tape you can choose to restore permisssions or not
as the case may be. If you choose not to restore permissions, you can
read any file that was restored. So if you have a backup the files are
not lost or inaccessible if you should have an AD disaster. You would
normally restore with permissions of course.

b) As we stated in the last post, is there no way to merge two domains? Or
merge the clients of an old one into a new one without blowing out their
profiles? None what so ever?

I didn't see the first post. It is possible to use the tools to move
users between Domains, but for a small number of users, it would
probably be easier to recreate them and their profiles.

c) In a small company, 10 stations, what should the tree be called? The
Wizard offers you choice a) YourCompany.local, or b) one could do
"yourcompany.com" ... The world wide DNS services are pointing
"yourcompany.com" to the router here -- but we're thinking of hosting the
website somewhere else later.

You can call the intermal Domain whatever you like. .local is fine.
Your company is like a fortress. You can get out and roam the outside
world, but the outside world is not allowed inside the castle. Same
with your DNS. The internal DNS is not known to the outside world DNS
and it only talks to the outside DNS if it needs to. The outside DNS
*never* talks to the inside DNS except to answer queries from the
inernal DNS. It just knows about the gateway. A postal worker from the
outside world delivers a letter to the gatekeeper. The gatekeeper
knows where to deliver the letter inside. That's like the outside
world accessing your web server which I presume is inside.

d) If you build the AD tree by hand, you can call it what you want -- but if
you let the wizard do it, it gets called something like "Your First
Site"...can that be renamed? Is this visible name important? Probably yes,
because it's somehow tied into the DNS at some point, no?

You can't build it by hand, but installing AD brings lots of tools.
Your site can be renamed, using the tools and the correct changes are
made to AD and DNS. However it doesn't really matter. My site is still
called "Default First Site" or whatever.

Replication
That part doesn't make sense to me at all. We certainly are not in a
position to purchase 2 servers -- if I understand your suggestion. Are you
telling me that there are no small offices running everything off of one
server? That that is considered bad practice? I'm talking about the
thousands of SOHO /small offices across the country. Are you really telling
me that every home network with rights has to have two servers or they have
to re-install their client profiles everytime they buy a new server to
upgrade? Serious?!?!

It is considered *best* practise to have two DC. That doesn't mean it
is always possible. I would certainly budget for one next year if you
can. It doesn't have to be a server class machine. A medium desktop
would do. If you can do it. Like I said, it's not always possible, and
there are many small offices that run only one server.

You do not have to reinstall anything when you buy a new server, if
you can add the new server to the AD Domain. You can add the server as
a member server, then promote it to a second Domain controller, then
demote the original DC to a member server, then remove it. There's a
bit more to it than that, but that's the core.

Scott is 100% right in what he says. If I make encouraging noises, it
doesn't mean that I disagree with him. It still would probably be best
to get some expert help. Consider. A small company's IT is its
backbone. If it should fail, the company might fail.

Cheers and good luck, once again.

Cliff

(MVP)

 

[Sky]

Dear Cliff (and Scott):

First, both, don't get me wrong: I would GLADLY, give the job back to a
consultant ;-) ...but "c'est la vie!" and I've got to muddle through it...

And both of you are being great helps. Really. Scott, I'm sorry my first
comments back were... a little gruff. (I've had some coffee since...)

DNS:
Internal DNS. Sounds like we've covered the basics here.
Gateway to External DNS.

That corrects one perception I had: I thought that we were all one big happy
worldwide DNS structure, just that it had to be Gatewayed to connect them.
What I think I now understand is that they have little to do with each
other...The internal DNS handles the names the computers in the office, and
for addresses beyond that range, gateways to the External DNS server. As for
the other way around the world cannot query the internal DNS? I mean not
possible to somehow navigate from home to (e-mail address removed)
(assuming firewall was off)... Not that I want to do this, but am interested
in knowing how that half works...


DHCP:
The DHCP ... Good to know that setting up in basic mode with no special
Options setup is ok...
Question: What services would you, personally, select in an office of 9
clients of Win2000. Two more on WinXP. Could you list out which ones I
should see as checked?


AD...and NTBackup
We had an NTBackup tape setup to do a full backup of D:\ (our data drive)
every week, and an incremental of changes every night.
We did not have one for C:\ (os & apps).
I mention this because, when we tried to restore from the tape after our
crash, the tape reported that it could not restore because it was unable to
recreate its catalog. We tried everything...but nothing. It succeeded with
the incrementals (because smaller size?) In the end, we were forced to copy
from the damaged hard-drive as much as we could get, but we've been left
with a whole bunch of excel files that are unable to be opened, and if so,
not saved. Either it's because, the files have SID's on them from clients
that no longer exist (? is this possible? Can this be cleared?) or simply
they are damaged. Not sure yet.

I mention this because at this point, we're very scared of NTBackup...didn't
deliver when we needed to restore...So to use it again, to backup our whole
server every night...You can understand our hesitation...

But you mentioned that you can check it once in a while. I understand that I
could restore a backup of D:\ to E:\ to check...but how do you restore an
Server's C:\ drive to another drive to check it? The files might all be
there, but...you don't actually know whether they 'work' together? Or did I
misunderstand? Actually, considering this, if a server crashed, and you
re-installed the OS, enough to run NTBackup, and restored...would that work?
Or would there be conflict/ dll hell as you tried to restore one OS onto an
existing OS?

Now. Assuming that I mention getting another cheap box to the boss, and I
don't get my head handed back on a spike, what are the steps?
a) Install new OS.
b) Install DNS. ? If so, what settings exactly?
c) Install DHCP? Is so, what settings?
e) Install AD? If so, what settings?
f) Anything else that is needs to be done?

You described my current topology as "root server in the root domain of a
one domain tree in a single tree forest!"....after that is done, does it
become "a secondary server in a root domain of a one domain tree in a single
forest?" (Just to verifying the lingo...)

and finally, if server 1 crashes, what's the procedure to switch to server
2? Vice versa, if server 2 crashes, nothing to do but resinstall, right? Do
I have to backup both at night? Or just one? And by the way -- doesn't
backing up all of C:\drive max out your tapes? And put a lot of wear on your
hard-drives as it opens/reads/saves every single file on your computer every
night?

Again, Scott, and Cliff -- thanks for all your help.
Sky

 

[Enkidu]

Hi Sky,

I'm quite happy to help out with the answer to your questions but this
is getting a bit looooong! I can understand your need for information
but, really, you'll need to get some education either through
self-teach or classroom. Anyway, please see inline for a few
answers/suggestions.

Cheers,

Cliff

Dear Cliff (and Scott):

First, both, don't get me wrong: I would GLADLY, give the job back to a
consultant ;-) ...but "c'est la vie!" and I've got to muddle through it...

And both of you are being great helps. Really. Scott, I'm sorry my first
comments back were... a little gruff. (I've had some coffee since...)

DNS:
Internal DNS. Sounds like we've covered the basics here.
Gateway to External DNS.

That corrects one perception I had: I thought that we were all one big happy
worldwide DNS structure, just that it had to be Gatewayed to connect them.
What I think I now understand is that they have little to do with each
other...The internal DNS handles the names the computers in the office, and
for addresses beyond that range, gateways to the External DNS server. As for
the other way around the world cannot query the internal DNS? I mean not
possible to somehow navigate from home to (e-mail address removed)
(assuming firewall was off)... Not that I want to do this, but am interested
in knowing how that half works...

Read up about private address ranges. Read up about NAT (Network
address translation. Briefly the only "external" address that is known
is your firewall's external address. The NAT box (gateway) keeps track
of who is talking to who, so is able to direct traffic to external
machines from internal machines. It can also direct traffic from
external machines to internal machines but a) this is a security risk,
and b) you have to specifically tell the NAT box to let the traffic
in. So it *is* possible, but it is complicated. By the way I think you
meant "computer3.my-company.com". The @ makes it an email address.

DHCP:
The DHCP ... Good to know that setting up in basic mode with no special
Options setup is ok...
Question: What services would you, personally, select in an office of 9
clients of Win2000. Two more on WinXP. Could you list out which ones I
should see as checked?

The DHCP options that are relevant are DNS server IPs and the gateway
IP address.

AD...and NTBackup
We had an NTBackup tape setup to do a full backup of D:\ (our data drive)
every week, and an incremental of changes every night.
We did not have one for C:\ (os & apps).
I mention this because, when we tried to restore from the tape after our
crash, the tape reported that it could not restore because it was unable to
recreate its catalog. We tried everything...but nothing. It succeeded with
the incrementals (because smaller size?) In the end, we were forced to copy
from the damaged hard-drive as much as we could get, but we've been left
with a whole bunch of excel files that are unable to be opened, and if so,
not saved. Either it's because, the files have SID's on them from clients
that no longer exist (? is this possible? Can this be cleared?) or simply
they are damaged. Not sure yet.

I mention this because at this point, we're very scared of NTBackup...didn't
deliver when we needed to restore...So to use it again, to backup our whole
server every night...You can understand our hesitation...

But you mentioned that you can check it once in a while. I understand that I
could restore a backup of D:\ to E:\ to check...but how do you restore an
Server's C:\ drive to another drive to check it? The files might all be
there, but...you don't actually know whether they 'work' together? Or did I
misunderstand? Actually, considering this, if a server crashed, and you
re-installed the OS, enough to run NTBackup, and restored...would that work?
Or would there be conflict/ dll hell as you tried to restore one OS onto an
existing OS?

I would suspect that the *tape* had a problem. You can make NTBackup
perform a validity check on what it has recorded, but that adds extra
time to the backup. Also you can restore a couple of files as a test.
Most tape drives need to have a cleaning tape run through them. I have
some tape DAT drives that I clean every week before the weekly backup.
I would not do a full restore as a test. Too time-consuming, too error
prone and you need to involve your live system. I would urge that you
make full backups every night. With a full/incremental system you
multiply the chances of a restore failure. Well, I prefer full backups
myself! A bad tape can be caused by using old tapes, having a dirty
drive or problems with the drive itself. Frequently used tapes stretch
and become unusable. NTBackup doesn't usually have problems.

Now. Assuming that I mention getting another cheap box to the boss, and I
don't get my head handed back on a spike, what are the steps?
a) Install new OS.
b) Install DNS. ? If so, what settings exactly?

You don't *have* to have DNS on all boxes. However it's a good idea!
Set the DNS as secondary to the other DNS. Set the box to use itself
for DNS.

c) Install DHCP? Is so, what settings?

Only one DHCP!! It's not a good idea to have two DHCP servers on one
network.

e) Install AD? If so, what settings?

Well, when you run dcpromo you get asked if you want to create a new
Domain (no) or new tree (no). All else is pretty obvious.

f) Anything else that is needs to be done?

Should be sweet after that. Some people make all DCs into Global
Catalog servers but that's optional (and opinions differ!) and is part
of the "advanced" course!

You described my current topology as "root server in the root domain of a
one domain tree in a single tree forest!"....after that is done, does it
become "a secondary server in a root domain of a one domain tree in a single
forest?" (Just to verifying the lingo...)
Exactly!

and finally, if server 1 crashes, what's the procedure to switch to server
2? Vice versa, if server 2 crashes, nothing to do but resinstall, right? Do
I have to backup both at night? Or just one? And by the way -- doesn't
backing up all of C:\drive max out your tapes? And put a lot of wear on your
hard-drives as it opens/reads/saves every single file on your computer every
night?

I understand that you've had a disaster. But they are fairly rare.
Really! However you are right to be prepared, but I'll not go into it
here. It's complicated and depends on your setup. Conceptually you are
right. If one fails restore the failed one. However there are
considerations. Are they both GC servers? What FSMO roles did the
failed one hold? (Look up FSMO! They are roles that each exist on a
single DC in a Domain as opposed to AD itself which is on all DCs)
There is heaps of documentation (which will boggle you at first) on
the Microsoft site and elsewhere.

Again, Scott, and Cliff -- thanks for all your help.
Sky

Your are welcome!

Cheers,

Cliff

 

[Sky]

Dear Cliff:

Wow! Thanks VERY VERY much for the time invested. I learnt a lot more in
your 2 emails than a stack of books. Not kidding. (Ever thought of writting
a book???)

I've certainly learnt a lot more in terms of the roles DHCP and DNS, and
NTBackup play, and what parts to read up more on.

I'm off to reinstall, setup as we discussed, reconnect the clients, clean
the Backup drive, run the fullbackup, and then go and try to convince them
to spend a little more dough on a second server.

Thanks for everything!