I dont build as many highend servers as i do desktop systems, but i am fairly competent with install and config. That being said, i have been working on this same problem for about 6 months or so, and i cant find a way around it.
situation: Client orders a server, i build and load it, and minor config stuff. intel entry level server boards, mirrored os drives (hardware when possible) and raid 5 data drives...
i load the OS, using floppy made from the intel disk for the F6 screen, basic stuff, this is 2k3 server std R2, and when everything is fine and loaded, i plug it into my companies network, and the whole network tanks, nobody can get out on the net, and every1 comes to me saying 'why the hll is the internet down'
from looking at my networking equipment and the activity on my switches when this problem is going on its only the port that the server and router is on that is experiencing heavy traffic.
specifics: ping www.google.com /t i get around 20-30ms responce, i plug the server in everything is fine, network usage on the server is between 0% and 0.01%, seconds later usage is constantly at 0.05% and ping is still getting 20-30 on test machine. about a minute later pings vary between 20 and 250, and a few minutes after that im getting 2000+ and im starting to drop packets. after about 10 minutes i have 100% packet loss (including pinging a raw ip to eliminate dns)
heres a twist: this is true on 3 completly different networks, my company's, my homes, and the other techs home, work has generic cisco equip and win 2000 servers, home has 2k3 servers and linksys router (dont ask) and everything is completly static, other guy idk what he has. heres another twist: same thing happens with SBS SP1, STD R2, and srv 2k3 enterprise (company VLP)
i have found an inconsistancy tho, (only applies to my enterprise vlp), everything works fine untill i get an RPC error with reminds me of the blaster worm, after it restarts is when everything goes to hll. to continue with this inconsistancy, i only get the RPC error when i install something like IIS or make it a DC.
im comparing services from all servers i have remote access to right now, ive allready disabled EVERYTHING except the 4 or 5 needed services and restarted with no luck
any ideas? maybe this is because its INTEL nic cards and mainboard? "seems" my tyan tiger 100 with a 3com nic worked fine.... because intel nic/mainboard is the only common link between all of these
pullin my hair out here, so is this y all the sys admins that i know are bald?
situation: Client orders a server, i build and load it, and minor config stuff. intel entry level server boards, mirrored os drives (hardware when possible) and raid 5 data drives...
i load the OS, using floppy made from the intel disk for the F6 screen, basic stuff, this is 2k3 server std R2, and when everything is fine and loaded, i plug it into my companies network, and the whole network tanks, nobody can get out on the net, and every1 comes to me saying 'why the hll is the internet down'
from looking at my networking equipment and the activity on my switches when this problem is going on its only the port that the server and router is on that is experiencing heavy traffic.
specifics: ping www.google.com /t i get around 20-30ms responce, i plug the server in everything is fine, network usage on the server is between 0% and 0.01%, seconds later usage is constantly at 0.05% and ping is still getting 20-30 on test machine. about a minute later pings vary between 20 and 250, and a few minutes after that im getting 2000+ and im starting to drop packets. after about 10 minutes i have 100% packet loss (including pinging a raw ip to eliminate dns)
heres a twist: this is true on 3 completly different networks, my company's, my homes, and the other techs home, work has generic cisco equip and win 2000 servers, home has 2k3 servers and linksys router (dont ask) and everything is completly static, other guy idk what he has. heres another twist: same thing happens with SBS SP1, STD R2, and srv 2k3 enterprise (company VLP)
i have found an inconsistancy tho, (only applies to my enterprise vlp), everything works fine untill i get an RPC error with reminds me of the blaster worm, after it restarts is when everything goes to hll. to continue with this inconsistancy, i only get the RPC error when i install something like IIS or make it a DC.
im comparing services from all servers i have remote access to right now, ive allready disabled EVERYTHING except the 4 or 5 needed services and restarted with no luck
any ideas? maybe this is because its INTEL nic cards and mainboard? "seems" my tyan tiger 100 with a 3com nic worked fine.... because intel nic/mainboard is the only common link between all of these
pullin my hair out here, so is this y all the sys admins that i know are bald?
Facebook
Twitter