LMAO, Why in the world did you go that route in the first place?
Seriously?
We have somewhere in the neighborhood of 1000 Macs, at least two thirds of those are managed with mcx/OS X Server software. We've been using XServes, XServe RAID's and there's even a couple of XSan setups in the company. Hell, most of the departments in the company are either Mac based or 100% Macs. Not to mention that the hardware has been rock solid.
We've been using Mac Pros rather than XServes for a while now, and maybe it's because we've never had it but LOM and a backup power supply are two things I haven't really missed much. We're a fair bit smaller than you though (200 in the UK, plus 6 international offices with about 20 each). Still, not being properly rackmountable and being aboslutely massive is kind of a pain in the sack.
ben0207 on
0
Options
mrt144King of the NumbernamesRegistered Userregular
LMAO, Why in the world did you go that route in the first place?
Seriously?
We have somewhere in the neighborhood of 1000 Macs, at least two thirds of those are managed with mcx/OS X Server software. We've been using XServes, XServe RAID's and there's even a couple of XSan setups in the company. Hell, most of the departments in the company are either Mac based or 100% Macs. Not to mention that the hardware has been rock solid.
That's a lot of Macs. I didn't mean to sound so flippant, it's just you're the first person I've talked to that wen the xserve route rather than like the other post above me mac pro route.
mrt144 on
0
Options
mrt144King of the NumbernamesRegistered Userregular
We got a new HP network printer several months ago (P3015) that has the stupidest driver. I haven't been able to find a bare bones print functionality only driver, which is annoying.
HP Universal Printing PCL 6? Or 5? It's super generic and doesn't work all that well with older models, maybe give that a shot.
I'm not a fan of the universal print driver. It works, but it seems slow and clunky, and it is glitchy with some of the not-terribly-old models I've tried it on.
Yeah, HP printers were fucking awesome back in the day of the Laserjet 4000 series. There are still a ton of those in active use. I'm pretty sure that after the nuclear apocalypse all the cockroaches will still be using Laserjet 4000s.
But each successive generation gets worse and worse. Don't even get me started on their all-in-ones. So much burning hatred.
The best new printers I've dealt with recently have been Brothers, BTW.
We have nothing but 4000s for most of our non MFPs. They're just solid simple pieces of equipment.
We have a bunch of Brothers for our new off site printers because they're pretty cost effective, have the functionality for what we need to do. It's just kinda odd that my fiancee has a brother sewing machine and that's how I've only known of them.
Posting over in the wow chat thread reminding me of the time I was a system administrator of a whole host of systems that were passworded with 'key.' One of which was a doctors office. That got hacked about 3 weeks after I started and I told them they should probably not use key anymore for a password. Bye bye patient records.
Also not HIPAA-compliant. They could have been sued if a patient found out.
Yup I let them know that quite verbally. I reported them to the state but not sure if anything ever really happened with that.
Luckily for them I don't think it was a hacker looking for information so much as a hacker looking to wreak havoc on a system and delete all the files.
bowen on
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
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
bowen on
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
Been having problems with our main server for a while at work because of the stupid 4GB memory limit of SBS 2003, ordered a replacement server today. A Dell R510, gonna be sticking SBS 2008 on there and maxing it out at 32GB of memory.
Spending a lovely £7k after getting a couple of different companies fighting over the price. (was originally £10k)
This is also after I just did a major infrastructure change, new server room, new server cabinet etc.. I love spending money that isn't mine.
EDIT:
I'm planning on getting a new file server next year, I was thinking of either getting one of the rack mountable readynas boxes or something from these guys. It'll be running either a Linux distro like CentOS or FreeBSD (maybe freenas). Anyone have any other suggestions? (like other people to buy from)
I'm aiming for 2u and as much storage capacity as possible. (hence 12 bays with 12 2TB drives in there setup in RAID6)
GrimReaper on
PSN | Steam
---
I've got a spare copy of Portal, if anyone wants it message me.
I have a VOIP problem and absolutely no idea how to fix it.
What kind of problems? I'll be setting up our office with VoIP in the next month or so and am curious about such problems, though I'm not sure how much I can help.
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
So, I should just be able to pull it out and pop it in while it's doing it's thing right? I am nervous, it's my first time, should I do it fast like a band-aid or slow and gentle?
bowen on
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
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
So, I should just be able to pull it out and pop it in while it's doing it's thing right? I am nervous, it's my first time, should I do it fast like a band-aid or slow and gentle?
This should probably be reported for awesome.
Also, nice and slow at first.. gently going in...
GrimReaper on
PSN | Steam
---
I've got a spare copy of Portal, if anyone wants it message me.
Well they use an onboard Intel storage manager. The problem is that they will throw up an error like "failed disk" and then you replace the disk and the next day it throws up the same error for a different port. It is unlikely that two disks fail hours apart from each other.
Then, the next day the controller throw an error reading "failed array" with no option to recover. It has happened a few times and it is always terrible. I tell people that these machines are flaky, and to consider things on them to be impermanent. I tell them to back up important things in Perforce or in their folder on the netapp, yet when the local raid fails it is always tears and "omg are you serious that stuff is gone??"
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
So, I should just be able to pull it out and pop it in while it's doing it's thing right? I am nervous, it's my first time, should I do it fast like a band-aid or slow and gentle?
This should probably be reported for awesome.
Also, nice and slow at first.. gently going in...
Try to do it at night if at all possible. That will help prevent problems performing.
In all seriousness, rebuilding a RAID array can slow down some of the server's other roles depending on the controller and the alignment of stars. If you don't want whinuses going "wahh, the server's slow!" then pop the new drive in around 4:30 in the afternoon.
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 remember an interesting article I read about observed multiple disk failures in striped RAID arrays. When building the array typically all the disks are bought at once and are identical, and thus are from the same manufacturing batch. They see the same duty cycles, and operate in the same operating conditions; therefore when one fails, it's not terribly rare for another in the set to fail (presumably not as rare as multiple concurrent failures occurring with a mixed set of disks). Or somesuch justification, cannot find the article now cause I'm lazy.
So I've got this 146GB 10K SAS drive. The array says it's hot-swappable. Can I actually hot swap this drive and have the RAID rebuild itself? Meaning if I pull out the third drive is the shit going to hit the fan? I've never had hotswappable hard drives before.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
So, I should just be able to pull it out and pop it in while it's doing it's thing right? I am nervous, it's my first time, should I do it fast like a band-aid or slow and gentle?
This should probably be reported for awesome.
Also, nice and slow at first.. gently going in...
Try to do it at night if at all possible. That will help prevent problems performing.
In all seriousness, rebuilding a RAID array can slow down some of the server's other roles depending on the controller and the alignment of stars. If you don't want whinuses going "wahh, the server's slow!" then pop the new drive in around 4:30 in the afternoon.
I did this, there wasn't actually any performance hit that was noticeable.
Also, Virtual Machines are way more headache than they're worth. The IT company before I came on board set all of their servers up as virtual machines spread across two servers. The speed of those machines was terrible mainly because of the ESX. The cost of upgrading the hardware/software was just too astronomical for them so I said, "Why don't we just move the RDP server to one of physical servers for the speed increase and then just keep the other server for whatever's left" (mostly DC/DNS/DHCP servers)
Wow the difference between the two is like the difference between a 486 and fucking i7.
bowen on
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
Also, Virtual Machines are way more headache than they're worth. The IT company before I came on board set all of their servers up as virtual machines spread across two servers. The speed of those machines was terrible mainly because of the ESX. The cost of upgrading the hardware/software was just too astronomical for them so I said, "Why don't we just move the RDP server to one of physical servers for the speed increase and then just keep the other server for whatever's left" (mostly DC/DNS/DHCP servers)
Wow the difference between the two is like the difference between a 486 and fucking i7.
Buh? That's odd... Last company I worked for had around 20 ESX hosts (running on relatively recent Dell hardware with a solid SAN and lots of RAM, mind you) and we could get 20+ guests running on those machines. Of course, we didn't put high disk throughput machines on there, but they all performed pretty well.
Also, Virtual Machines are way more headache than they're worth. The IT company before I came on board set all of their servers up as virtual machines spread across two servers. The speed of those machines was terrible mainly because of the ESX. The cost of upgrading the hardware/software was just too astronomical for them so I said, "Why don't we just move the RDP server to one of physical servers for the speed increase and then just keep the other server for whatever's left" (mostly DC/DNS/DHCP servers)
Wow the difference between the two is like the difference between a 486 and fucking i7.
Buh? That's odd... Last company I worked for had around 20 ESX hosts (running on relatively recent Dell hardware with a solid SAN and lots of RAM, mind you) and we could get 20+ guests running on those machines. Of course, we didn't put high disk throughput machines on there, but they all performed pretty well.
Sounds more like to me that they didn't configure them properly, if an ESX vm is running slow then the people who configured it did a shit poor job. They should also have used hardware that is capable of AMD-V or Intel VT-X.
GrimReaper on
PSN | Steam
---
I've got a spare copy of Portal, if anyone wants it message me.
0
Options
Apothe0sisHave you ever questioned the nature of your reality?Registered Userregular
edited November 2010
Absolutely. Virtualisation is hugely beneficial and properly configured and on appropriate hardware the performance impact is if not negligible small enough to not be a significant factor.
Yeah and fuck if I was going to do it. Much easier to have 2 physical hardware machines doing the same task as about 10 vm machines. I'm sure I'll probably reverse that once I have to reload those machines. But I'm wagering a guess it had something to do with it being on a shitty server, on a shittily designed network with a shitty NAS. Yes, a NAS.
I love those supermicro boards, they so powerful!
bowen on
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
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Yeah and fuck if I was going to do it. Much easier to have 2 physical hardware machines doing the same task as about 10 vm machines. I'm sure I'll probably reverse that once I have to reload those machines. But I'm wagering a guess it had something to do with it being on a shitty server, on a shittily designed network with a shitty NAS. Yes, a NAS.
I tried a NAS (TeraStore I had lying around) once just to experiment with some VM migration between hosts, it was easily the worst thing performance-wise I'd ever used. On the flip-side, I've got an old dual-Xeon (P4 era) box with 5x Ultra320 drives running FreeNAS as an iSCSI target, it's surprisingly quick.
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Anybody know an easy fix?
I do not, but I certainly sympathize.
Roaming profiles are evil, especially when people start keeping a gig of junk in their My Documents folder and then wonder why it takes forever to log on to a new computer.
chamberlain on
0
Options
lwt1973King of ThievesSyndicationRegistered Userregular
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Anybody know an easy fix?
What are the clients running? XP, Vista, 7?
lwt1973 on
"He's sulking in his tent like Achilles! It's the Iliad?...from Homer?! READ A BOOK!!" -Handy
Well they use an onboard Intel storage manager. The problem is that they will throw up an error like "failed disk" and then you replace the disk and the next day it throws up the same error for a different port. It is unlikely that two disks fail hours apart from each other.
Then, the next day the controller throw an error reading "failed array" with no option to recover. It has happened a few times and it is always terrible. I tell people that these machines are flaky, and to consider things on them to be impermanent. I tell them to back up important things in Perforce or in their folder on the netapp, yet when the local raid fails it is always tears and "omg are you serious that stuff is gone??"
RAID/server guy next office over says we had a similar problem and it was the RAID controller being crap.
I have a VOIP problem and absolutely no idea how to fix it.
What kind of problems? I'll be setting up our office with VoIP in the next month or so and am curious about such problems, though I'm not sure how much I can help.
We have a server at our main office that hosts config files for the phones. The config basically just tells the phones where the Acme server is and how to contact it when making phone calls.
If you're setting up VOIP on a network with Cisco equipment, then your master switch apparently needs an Interface Vlan with an IP from the DHCP pool created for the VOIP phones.
Dunno why, but the moment I created one the ARP table populated with the phones and they started hitting the config server.
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Anybody know an easy fix?
I do not, but I certainly sympathize.
Roaming profiles are evil, especially when people start keeping a gig of junk in their My Documents folder and then wonder why it takes forever to log on to a new computer.
re-direct their my documents through group policy. makes it so that the my docs folder acts pretty much like a network share (we do this for desktops too).
upside: cuts down on log on/off times considerably
downside: if the network craps they're cut off from it
i think there are more re-direct options in 2008 (my docs and the desktop are the only options in 2003, i think)... well... i hope there are; because i really want to quit using roaming profiles here. spend so much time troubleshooting/rebuilding profiles, and it's gotta be one of the most tedious things.
also, @smarthero, anything about the profile migration getting logged in the event logs on the local machines?
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Anybody know an easy fix?
What are the clients running? XP, Vista, 7?
XP on most machines, one or two are running 7 (slowly rolling out new workstations to staff)
@tyrant Haven't even looked. I'm only our 'part-time IT guy' (which is really dumb but a whole different discussion) and my regular work is keeping me too busy to dig too deep into the issue.
re-direct their my documents through group policy. makes it so that the my docs folder acts pretty much like a network share (we do this for desktops too).
upside: cuts down on log on/off times considerably
downside: if the network craps they're cut off from it
i think there are more re-direct options in 2008 (my docs and the desktop are the only options in 2003, i think)... well... i hope there are; because i really want to quit using roaming profiles here. spend so much time troubleshooting/rebuilding profiles, and it's gotta be one of the most tedious things.
also, @smarthero, anything about the profile migration getting logged in the event logs on the local machines?
I agree with this post.
Hero, I've encountered that problem before with roaming profiles. Sometimes it's because the profile isn't getting properly closed when the system is powered down, sometimes it's because of folder security on the profile directory, sometimes it's because Windows just decided to piss in your cornflakes that day.
Fuck roaming profiles. Very few organizations really have any need for them, and those organizations really should be looking at desktop virtualization instead.
User shell folder redirection all the way.
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 am in no way truly qualified to be our network/sys admin it was hoisted upon me with no training and even less documentation by our full-time predecessor (cursed be his name)
5AM/Pre-office hours speed test for a customer. Yippee. He's like "I want to speedteset the exact interface that I'm connected too". Can't do that, just to a server that's on the network.
But, hey, work from home OT. Can't beat that.
Sheep on
0
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
edited November 2010
I have a user unable to RDP into her desktop machine. I have learned that the shortcut she's using is pointed to the WAN address (router has a port forwarding rule) and the laptop she's trying to connect from is on the same network.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I have a user unable to RDP into her desktop machine. I have learned that the shortcut she's using is pointed to the WAN address (router has a port forwarding rule) and the laptop she's trying to connect from is on the same network.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I don't think it would be a firewall issue unless there were access control lists setup to prohibit RDP'ing from machines within the same network.
Sheep on
0
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
I have a user unable to RDP into her desktop machine. I have learned that the shortcut she's using is pointed to the WAN address (router has a port forwarding rule) and the laptop she's trying to connect from is on the same network.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I don't think it would be a firewall issue unless there were access control lists setup to prohibit RDP'ing from machines within the same network.
I wonder what the issue is then? I tested it from outside with the WAN address and it works fine. The user seems satisfied with the dual shortcut solution, luckily.
I have a user unable to RDP into her desktop machine. I have learned that the shortcut she's using is pointed to the WAN address (router has a port forwarding rule) and the laptop she's trying to connect from is on the same network.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I don't think it would be a firewall issue unless there were access control lists setup to prohibit RDP'ing from machines within the same network.
I wonder what the issue is then? I tested it from outside with the WAN address and it works fine. The user seems satisfied with the dual shortcut solution, luckily.
What port is being forwarded? You may have to change the RDP listening port on the destination PC.
chamberlain on
0
Options
TL DRNot at all confident in his reflexive opinions of thingsRegistered Userregular
I have a user unable to RDP into her desktop machine. I have learned that the shortcut she's using is pointed to the WAN address (router has a port forwarding rule) and the laptop she's trying to connect from is on the same network.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I don't think it would be a firewall issue unless there were access control lists setup to prohibit RDP'ing from machines within the same network.
I wonder what the issue is then? I tested it from outside with the WAN address and it works fine. The user seems satisfied with the dual shortcut solution, luckily.
What port is being forwarded? You may have to change the RDP listening port on the destination PC.
It's a simple setup on a small domain; the router forwards traffic on port 3389 to her workstation, and 3389 is the default listening port for RDP.
On recent versions of windows server there's the terminal services gateway service that could be used to simplify a lot of the firewall messing about, depends whether they have a server capable of running it though.
On SBS 2003 and 2008 it's pretty much entirely pre-configured for you and ready to go. Login to website, click on your workstations and bam you've logged into your desktop via rdp.
GrimReaper on
PSN | Steam
---
I've got a spare copy of Portal, if anyone wants it message me.
Posts
We've been using Mac Pros rather than XServes for a while now, and maybe it's because we've never had it but LOM and a backup power supply are two things I haven't really missed much. We're a fair bit smaller than you though (200 in the UK, plus 6 international offices with about 20 each). Still, not being properly rackmountable and being aboslutely massive is kind of a pain in the sack.
That's a lot of Macs. I didn't mean to sound so flippant, it's just you're the first person I've talked to that wen the xserve route rather than like the other post above me mac pro route.
We have nothing but 4000s for most of our non MFPs. They're just solid simple pieces of equipment.
We have a bunch of Brothers for our new off site printers because they're pretty cost effective, have the functionality for what we need to do. It's just kinda odd that my fiancee has a brother sewing machine and that's how I've only known of them.
Yup I let them know that quite verbally. I reported them to the state but not sure if anything ever really happened with that.
Luckily for them I don't think it was a hacker looking for information so much as a hacker looking to wreak havoc on a system and delete all the files.
Spending a lovely £7k after getting a couple of different companies fighting over the price. (was originally £10k)
This is also after I just did a major infrastructure change, new server room, new server cabinet etc.. I love spending money that isn't mine.
EDIT:
I'm planning on getting a new file server next year, I was thinking of either getting one of the rack mountable readynas boxes or something from these guys. It'll be running either a Linux distro like CentOS or FreeBSD (maybe freenas). Anyone have any other suggestions? (like other people to buy from)
I'm aiming for 2u and as much storage capacity as possible. (hence 12 bays with 12 2TB drives in there setup in RAID6)
---
I've got a spare copy of Portal, if anyone wants it message me.
What kind of problems? I'll be setting up our office with VoIP in the next month or so and am curious about such problems, though I'm not sure how much I can help.
Assuming the array isn't RAID 0 and the controller supports it, it should start rebuilding when you swap in the new drive.
Raid 5 - Smart Array P400 Controller (SAS) Hotswappable it says.
So, I should just be able to pull it out and pop it in while it's doing it's thing right? I am nervous, it's my first time, should I do it fast like a band-aid or slow and gentle?
This should probably be reported for awesome.
Also, nice and slow at first.. gently going in...
---
I've got a spare copy of Portal, if anyone wants it message me.
Then, the next day the controller throw an error reading "failed array" with no option to recover. It has happened a few times and it is always terrible. I tell people that these machines are flaky, and to consider things on them to be impermanent. I tell them to back up important things in Perforce or in their folder on the netapp, yet when the local raid fails it is always tears and "omg are you serious that stuff is gone??"
Try to do it at night if at all possible. That will help prevent problems performing.
the "no true scotch man" fallacy.
I did this, there wasn't actually any performance hit that was noticeable.
Also, Virtual Machines are way more headache than they're worth. The IT company before I came on board set all of their servers up as virtual machines spread across two servers. The speed of those machines was terrible mainly because of the ESX. The cost of upgrading the hardware/software was just too astronomical for them so I said, "Why don't we just move the RDP server to one of physical servers for the speed increase and then just keep the other server for whatever's left" (mostly DC/DNS/DHCP servers)
Wow the difference between the two is like the difference between a 486 and fucking i7.
Buh? That's odd... Last company I worked for had around 20 ESX hosts (running on relatively recent Dell hardware with a solid SAN and lots of RAM, mind you) and we could get 20+ guests running on those machines. Of course, we didn't put high disk throughput machines on there, but they all performed pretty well.
Sounds more like to me that they didn't configure them properly, if an ESX vm is running slow then the people who configured it did a shit poor job. They should also have used hardware that is capable of AMD-V or Intel VT-X.
---
I've got a spare copy of Portal, if anyone wants it message me.
I love those supermicro boards, they so powerful!
We've got roaming profiles set up on Windows Server 2003 for our users, so they can hop on to any workstation with their account and have all their docs and whatnot.
Recently, however, there's been a weird issue where if they delete files on their local machine a log out, when they log back in the files are 'undeleted' as they put it and now there's a second copy of the file.
What I'm guessing is that their roaming profiles aren't syncing back to the server-side copy of the profile properly, so it's restoring what they've deleted each time they log in.
Anybody know an easy fix?
I tried a NAS (TeraStore I had lying around) once just to experiment with some VM migration between hosts, it was easily the worst thing performance-wise I'd ever used. On the flip-side, I've got an old dual-Xeon (P4 era) box with 5x Ultra320 drives running FreeNAS as an iSCSI target, it's surprisingly quick.
I do not, but I certainly sympathize.
Roaming profiles are evil, especially when people start keeping a gig of junk in their My Documents folder and then wonder why it takes forever to log on to a new computer.
What are the clients running? XP, Vista, 7?
RAID/server guy next office over says we had a similar problem and it was the RAID controller being crap.
Apparently RAID sucks.
We have a server at our main office that hosts config files for the phones. The config basically just tells the phones where the Acme server is and how to contact it when making phone calls.
If you're setting up VOIP on a network with Cisco equipment, then your master switch apparently needs an Interface Vlan with an IP from the DHCP pool created for the VOIP phones.
Dunno why, but the moment I created one the ARP table populated with the phones and they started hitting the config server.
re-direct their my documents through group policy. makes it so that the my docs folder acts pretty much like a network share (we do this for desktops too).
upside: cuts down on log on/off times considerably
downside: if the network craps they're cut off from it
i think there are more re-direct options in 2008 (my docs and the desktop are the only options in 2003, i think)... well... i hope there are; because i really want to quit using roaming profiles here. spend so much time troubleshooting/rebuilding profiles, and it's gotta be one of the most tedious things.
also, @smarthero, anything about the profile migration getting logged in the event logs on the local machines?
XP on most machines, one or two are running 7 (slowly rolling out new workstations to staff)
@tyrant Haven't even looked. I'm only our 'part-time IT guy' (which is really dumb but a whole different discussion) and my regular work is keeping me too busy to dig too deep into the issue.
I agree with this post.
Hero, I've encountered that problem before with roaming profiles. Sometimes it's because the profile isn't getting properly closed when the system is powered down, sometimes it's because of folder security on the profile directory, sometimes it's because Windows just decided to piss in your cornflakes that day.
Fuck roaming profiles. Very few organizations really have any need for them, and those organizations really should be looking at desktop virtualization instead.
User shell folder redirection all the way.
the "no true scotch man" fallacy.
how do I do that instead
i am in no way truly qualified to be our network/sys admin it was hoisted upon me with no training and even less documentation by our full-time predecessor (cursed be his name)
But, hey, work from home OT. Can't beat that.
I imagine it's a conflict of some sort at the firewall, trying to pass traffic out and back in on the same address/port? Perhaps a second shortcut with her workstation's LAN address would be the way to go here.
I don't think it would be a firewall issue unless there were access control lists setup to prohibit RDP'ing from machines within the same network.
I wonder what the issue is then? I tested it from outside with the WAN address and it works fine. The user seems satisfied with the dual shortcut solution, luckily.
What port is being forwarded? You may have to change the RDP listening port on the destination PC.
It's a simple setup on a small domain; the router forwards traffic on port 3389 to her workstation, and 3389 is the default listening port for RDP.
On SBS 2003 and 2008 it's pretty much entirely pre-configured for you and ready to go. Login to website, click on your workstations and bam you've logged into your desktop via rdp.
---
I've got a spare copy of Portal, if anyone wants it message me.