We've been investigating software and appliances that would allow us a central, web based login to manage access and users to all the servers with IPMI cards. Does anyone use anything other than appliances from avocent or raritan?
I have a question maybe someone can help with. Is it possible to set the IP address on a IPMI card before a system ships? That way when the customer gets it, it already has the IP address set to whatever was agreed on and they can log right in using that IP address?
The issue we have is that server1 (web) has the load very high, compared to server2 (mysql). Sever2 stays at 0.5 during peak time.
I must mention that we have very heavy traffic hits only for 30min/day, then things calm down.
I simple diagram would be: [USERS] <= NIC1 => [SERVER1 WEB] <= NIC2 => [SERVER2 MYSQL]
So I think we can install a load balancer and upgrade the NIC cards, in order to solve this problem.
What I thought is this:
In server1 (web) install 2 NICs, one facing the users and one facing the server2 (mysql). Currently we have a 100MB NIC card into each server. The 95th is way to high at peak time, over 500MB so I was thinking on installing a 1GB NIC facing the server2 (mysql).
My question is: What is the best way to install a second NIC into an existing configuration? Do I need a Cisco 2970 switch?
I was wondering if there's any web hosts that offer 3D graphics cards in their hosted dedicated servers? they don't have to be that great of 3D graphics cards just better graphics than the normal video cards that they usually put in them?
As I am going to launch unique product connected with webhosting in the near future, I would like to learn more about the tech aspects of webhosting. By tech aspects I mean for example how to setup nameservers - how they work and how to setup them, SSL, how to setup ftp etc., so, basically I am interested in all the things connected to servers.
Could you please recommend me some guides or books related to webhosting and server management. I am newbie to servers and hosting, so I would like to learn it from the ground. I am searching for some book for beginners, but I cannot find any.
Knownhost has Servers in Texas and in California. You get double the bandwidth (with no price increase) if you take the California servers.
I live in Vancouver Canada which is on the west coast above California. So when I ping/traceroute Texas vs. California, the California servers do better for me sitting at my desktop.
My question is this...
What is better for people who will be hitting my pages? If most of my visitors will be scattered throughout the U.S. is it a lot better to have my pages hosted centrally in Texas, or is there only going to be marginal difference.
Would you take double the bandwidth on the west coast or would you prefer your servers located centrally in the U.S.?
I am really not sure if the location is effecting to speed. I am chooing a provider for colo'ing few servers. I know each region has very good providers (in both services and network, from reviews on WHT), but - from the west coast to central, the latency difference is around 30 - 40ms - from the central to the east coast, the latency difference is around 30ms
so, if I go with a provider in west coast, it really saves me 60-70ms latency from my area. However, someone likes gnax in east coast using Route Science, that's advertised to help to get the better routes.
I dont see any in west coast using RS or FCP. So, should I go - with a "normal" provider (I mean network without RS or FCP equipments) in west coast, - or with a provider with RS/FCP in any place?
Not sure on industry standard here, but I am now curious about this issue.
I placed an order with a company, paid for a setup, first month etc. On receiving the order, I was given the wrong product. Which was fixed for me within a few days. I was happy to even pay for this which the host declined.
Fast forward a week or so. Still in the first month, and issues are still not resolved.
I've now removed credit card information from the billing system (yes I know, shouldn't do it) with good reason. I expected to have to pay another month, and that was fine.
Anyway, month ends and I am charged on a credit card that was removed from the billing system. When confronted about it, I was told standard practice. I think this is highly unethical to hold details after they had been removed.
Is this indeed standard practice in this industry? Do a majority of other hosts do the same thing?
Now again, I am only to happy to pay another month, just not from a credit that was removed, which is why it was removed. I would've sent a check, or done a deposit. I do understand that fraud, etc is very high in this industry. I have a host or two not give me service because of the email address I used during sign up, but when changed, I received service.
i have several servers here but only basic knowledge about raid
My understanding of raid is like this
I have a server with 4 disk on it and if i put it in raid 1 and want to install an OS i can only see 2 of the disks right? (thats how it works on my ibm server)
But with my other servers if i put them in raid 1 and want to install OS i can still see all 4 disks thats not right is it?
anyone that can recommend a pci raid card for 4 sata2 disks?
What are the benefits of IPMI with KVM Over LAN? Is it more of a benefit for the datacenter? Or does it have benefits for someone that has collocated there server a far distance away?
This may appear to be a stupid question, but I just wonder if the Supermicro IPMI cards work only with Supermicro hardware (or if it works also with other hardware, but without warrantly)?
In case does it exist some hardware independant, or something similar?
to setup a Windows server that can support the failure of the switch that comes before it in the network diagram. The idea would be that if the 15A circuit the switch was on failed, or the switch just died, the server would still be online.
I already have two switches in spanning tree just before the server. Is there a way to assign a single IP to the server on both its network cards and connect each card to a different (spanning tree'd) switch?
I am setting up an internal network for management only. So, assign an IP for the second NIC, and activate it, but it seems not working. I have tried this
#ifconfig eth1 192.168.2.14 netmask 255.255.255.0 broadcast 192.168.2.255 #ifconfig eth1 up
checking dmesg, it shows the NIC is up # dmesg ADDRCONF(NETDEV_UP): eth1: link is not ready tg3: eth1: Link is up at 100 Mbps, full duplex. tg3: eth1: Flow control is off for TX and off for RX.
checking routing table seeing the 192.168.2.x routed through eth1 # route -e 192.168.2.0 * 255.255.255.0 U 0 0 0 eth1 192.168.1.0 * 255.255.255.0 U 0 0 0 eth0 169.254.0.0 * 255.255.0.0 U 0 0 0 eth1 default reserve1.somename 0.0.0.0 UG 0 0 0 eth0
Repeat those steps for other boxes, but when pinging, all return errors # ping 192.168.2.20 PING 192.168.2.20 (192.168.2.20) 56(84) bytes of data. From 192.168.2.14 icmp_seq=1 Destination Host Unreachable From 192.168.2.14 icmp_seq=2 Destination Host Unreachable From 192.168.2.14 icmp_seq=3 Destination Host Unreachable
I'm in the market for a new server for colocation, and to save myself some of the headaches I've had with my current one, I'd like it to have a good remote management system, specifically one that operates independently of the operating system, or ideally one that can be accessed even if the machine is powered off. What exactly should I be looking for? I've read a lot about IPMI and Intel vPro and AMT, and I'm not quite clear on how they all fit together.
At the very least, I need something that will let me power-cycle the machine and remotely access its serial console. All the hardware environment monitoring things in IPMI sound useful, but not essential for my purposes. Anyway, it sounds like vPro is part of many (or all?) current Core 2 processors, or am I mistaken? I'm looking mainly at machines with Core 2 Duo or Quad, so can I expect they'd have that built-in, or would I need to specifically request it or look for some particular chipset or CPU?
I am having alot of difficulty trying to setup supermicro IPMI.
I am using motherboard PDSMI+.
Ok, so this is what I did:
1. I flash the IPMI card with the cd-rom that is included. I set the correct IP and MAC address.
2. I then install OpenIPMI-tools
I was able to successfully issue the command locally: ipmitool sensor
but, when I try to connect remotely I always get a connection failed. I tried everything I know, which is nothing and got no progress.
I set the ipaddres and gateway by using this commands: ipmitool -I open lan set 1 ipaddr [IPADDR] ipmitool -I open lan set 1 defgw ipaddr [GW IPADDR] ipmitool -I open lan set 1 netmask [NETMASK] ipmitool -I open lan set 1 access on
Is it possible to use a Central MailServer Proxy (eg. Dovecot Director/Proxy or Nginx). We have many Plesk Server. All with Postfix and Dovecot.... On each Server is a SSL Cert for MailService but we want ONE Host... mail.example.com with one SSL Cert.
How is it possible to Sync the User from all Hosts with a MailProxy? Only so customers can be transferred from server to server.
I have a bunch of servers with IPMI cards with KVMoIP capability and I know how to configure IPs and everything using IPMICFG.EXE while the system boots.
Now, I am facing a dilemma as to how to create the IP space.
When I create a user I specify the user has no access to change the network (IP) info.
But say the user has windows server, are they able to change the IP from there using IPMICFG.EXE?
This will determine how I create my VLANs.
If the client with USER access is able to change the IP settings using IPMICFG.exe then I have to create a /30 VLAN for each KVMoIP. I dont want to do this cause 3 IPs are wasted.
If the user will not be able to execute IPMICFG.EXE then I can just create one big VLAN and assign one IP per IPMI card with the gateway being the same for the entire big VLAN.
I have here 3 servers of Tyan with M3291 IPMI boards installed. I hooked it up as per instructions of Tyan, but am unable to get it to work.
I have tried contacting Tyan about this, but after a few messages they stopped responding to my questions. I think they are convinced it is working with me....
The setup is; - Ubuntu 6.1 x86_64 - System BIOS is the original version - IPMI card is flashed with the correct firmware - The network port which is shared with the IPMI card has a different IP address then the IPMI card.
Settings IPMI card: box:~# ./uh8l -d BMC Utility on Linux For M3291 version 0.9.1 Copyright 2003-2006 TYAN Computer Corp. IPMI Ver: 2.0 Firmware Rev: 0.11 Vendor ID: 6653 Pdt ID: 5161 DHCP Disable BMC IP Address: 10.0.1.102 MAC Address: 00:E0:81:42:4C:51 Net Mask: 255.255.255.0 Gate Way: 10.0.1.253 Broadcast ARP: Enable
This server has on the IPMI shared port the address 10.0.1.2 ... When I try to ping the IPMI board from the server itself or from a different server I get an incomplete address in the ARP table; ? (10.0.1.102) at <incomplete> on eth2
(Also after pinging it for 1 minute...)
If I use the IPMI utilities I am unable to contact the board and do anything with the server.... I tried several settings of the IPMI board, including DHCP / ARP on and off...
Does anyone have this combination working, and how did you do this?
After upgrading from V11 to V12 : file not found in the central Frame
After upgrading from V11 to V12 version, I enter my credentials to login and when I get the main page, I have in the main frame : "File not found". : See screenshot
I was able to login from IPMI View just fine, and everything works as expected except "Text Console ". I was able to hit the start button and get the black screen on IPMI View, but that was just about it.
Only a cursor blinking without texts.
Is there something else that I should do in order to get text console working? I've tried all baud rates.