There is no doubt in my mind I'm going to Houston. The prices they have..it's unreal. Take a look:
[url]
And that's just one. The real fun starts when they start beating each other's quotes.
Ok..on to what need here:
If I have a cage at Level 3..is there a way to estimate routing/switching needs on a per-rack basis? In other words..can you estimate: Ok 1 full rack...say...30 servers..and you're bringing in one of those 100mbps eth drops...to start.
and you want capacity for..say..three more racks from the getgo.
Is there software for this? Or can you guesstimate on ip use-how many ips you think you'll need routing for?
how much power the average hard drive would use? (ie a 1tb drive). I was looking on WDC's website and it said 7.5 watts peak durring writing and reading etc, so assuming thats at 110 voltage, would it be safe to assume that 12 drives @ 7.5 watts each would make a total of .81 amps? I am going off of the equation that Amps = Watts/Volts. So amps = 90/110
I am looking at some rackmount chassis and most of them come with power supplies included. Most of the 1U that I look at have 250-300watt power supplies.
The way I see it is 2 processors @ 130watts is 260watts right there. Thats not including hard drives, and the motherboard. Am I calculating this wrong because these power supplies seem too small? Also is there any place that lists motherboard and hard drive watts or is there a general industry rule of thumb? Does the number of ram sticks in a motherboard affect how many watts a motherboard uses. Thanks.
Proliant ML570 4x 2.0GHz Xeon MP I'm gonna add 3GB of PC1600 ECC RAM for a total of 4GB I'd like to have the following hard drive config:
2x18 GB 15k RPM in RAID1 for OS (CentOS 5.0) 3x36+ GB 10k RPM in RAID5 for data
However that drive config may suck a lot of power.
The server has 3x600w PSUs, however, if I keep it under 6 drives, I can set up the PSUs in a 1+1 configuration so I would only HAVE to have 1 PSU connected.
The idea would be that I could lose any 15A circuit, switch, firewall or service provider and still remain up. Full routes for the routers. Spanning tree on all the switches.
We have the OpenBSD routers up and running in our lab, I'd say they are 90% ready, but we're having internal routing issues when it comes time to fail-over. I know we'd have an easier time if we used just one router with a NIC per provider, but I was hoping for more redundancy.
I do not know if this is the right forum to post on, or if this is even the right site since this is not technically "Web Hosting" related, however it is the largest technical community I know of.
I have been setting up small networks in various branches of a medical company. I am using a linksys wired router connected to a static IP through a SDSL modem. The DHCP is disabled, and the two computers and the print server have also been assigned static IPs.
The problem I am having is that the linksys router will work fine, and then go offline.
The linksys router will no longer be pingable internally. Unplugging the router from the power briefly then plugging it back in resolves the problem temporarily. Sometimes it will become unresponsive again within 30 seconds to a minute, however other times it will go days without this problem occuring.
I thought it might be a defective router, however I took it back and got a replacement, however this is continuing to occur, so it must be something else. The two computers are Windows XP with static IPs set to 10.88.102.xxx as well as the print server. and the subnet mask is 255.255.255.0. The gateway has the IP 10.88.102.1. The mask is set as the same on the router side. The DNS server IPs are the ones given by the ISP and are defined in both the client TCP properties and on the router.
I figured since the network was so small, it would make sense to just buy a cheap linksys router, however the response I got from someone at experts-exchange says that this is just the MO of the crappy residential end of linksys. This surprises me since linksys is a division of cisco.
I have a server with several sites hosted on it. Randomly, a site user will find that they can no longer connect to any of the sites on the server - but they can connect to the rest of the Internet. When this happens, the sites themselves are fine and everyone else can connect without any problems.
The only way the affected user can view sites on the server again is to reboot their router - at which point everything is fine.
why this would happen? I have a CentOS (Red Hat Enterprise) Linux server with DirectAdmin, Apache, PHP/MySQL - the usual. I'm thinking that there must be some kind of network setting which is allowing this to happen, but I can't for the life of me work out what it is.
My roommate ran into this problem a few weeks ago. He could not connect to a webserver. Turns out it is the way that some of the ports are being handled when passive FTP is used.
I was wondering if a lot of web hosting companies are seeing some complaints of (passive) FTP not working properly? And who are you blaming?
He had to install some firmware updates to the Linksys router as well but then that started to crash the Internet connection (something we cannot have). So the router update was uninstalled.
He had to do quite a bit and finally started to use the control panel to upload. He has not tried FileZilla yet but I tried it on my system to connect to his server and it failed.
Please give me the difference. Colo in carrier hotel, we can choose our preferred network provider, but should we do that if we cannot have our own tech in datacenter? How about the supporting service from carrier hotel? Just general question, cause I dont address exactly which facility.
And the second would be more expensive? Saying the same number of rack, amount of bandwidth... Who is providing IP addresses then?
I've been out of the game for a while and now looking at colo prices each server is only allocated 0.5a on most plans.
I was looking at purchasing a 1u HP DL160 dual quad core system with at least dual sata raided drives
My question is what kind of amps would a system like this pull? and how much do data centers typically charge for that additional power if it needed over the .5
if anyone knew hosts (other then FDC in USA) that offers 1U-4U/midtower colocation with 100Mbit or greater uplink, with atleast 2TB of transfer. With IRC allowed.
My budget is ~$100, I'm fairly sure it can be done as I saw FDC had one for $79, and it would be fine though I was hoping if anyone knew any others.
Also, Giga-International has what I need, are they reliable?
colo of a 1u server that would need 500gig per month of b/w and I perferably would like to find a DC in NY/northern NJ or southern CT although in my search I seem to be getting price quotes of $100+ per month which I think is insane when I see dedicateds with more b/w for the same amount or even less.
I know most will say just get a dedicated server somewhere but my requirments are that I need a server with a lot of ram and at least a dual cpu and dedicate's with a dual cpu and 2gigs or more of ram seem to be much more thenmy budget.
So any suggestions for a $60 or less per month colo space with 500gig of b/w or a dedicated server provider offering a server in that price range with dual cpu and 2gig ddr ram and 80gig hdd?
By the way I looked at ezzi.net which has a $49 deal on a dedicated server but no option to have one with a dual cpu:
I have had experience with reselling hosting using HostGator seller. My job was basically to run my site and get customers, set up price plans etc on WHM. I sold that company a while back. I am ready to have another shot at hosting, but this time I want to use my own servers...
I have found a great site, which has customisations etc. on servers (http://www.cybertronpc.com), but they don't ship to UK. (If I'm using colo is US, could I get it shipped straight to them?)
My main question is, is colo needed? If I am going to setup this company the Data-Center is not going to be local, either London or in US. So there is no chance of me going down unless it is to pick up my server . So I think I'm going to need a maintaned service I think. The server will be used for clients data (shared hosting). So what services will I need in terms of security and stability? What am I looking for in a colo service? What about back-ups of data on the server? Is that my responsibility or can it be bought as a service? Ok... now I'm guessing that I'm going to need colo...
What are common problems etc. with servers? Am I going to need virtual IP connection for maintainance... Is it best I pay for this or a service operator? Any recommendations on where to have the colo (i'm based in UK) and why? What can I do when it is time for me to request my server back from UK?
When changing colo services is there any way of avoiding down-time?
currently I have managed several dedicated servers. I plan to colo it with 1 rack at a datacentre. So, before I buy hardware or software, I need some helpful info/guide.
- I need my server can be monitored, reboot remotely. What kind of hardware is require and please suggest some models.
- Which software is suitable for billing, monitoring. Please suggest any software come with good API since I plan to develop own small control panel later.
What is the best indication for a web hosting company to move from dedicated server to colocation?
I have several low end dedis and im thinking of buying an enterprise class server with lots of diskspace (raid 5, dual power, ecc, etc.), have it colocated and move all the accounts to that server. I would be saving in the long run but kinda turned-off due to all eggs in one (enterprise-class) basket dilema. I woud be saving on server management cost too because I'm signing up one machine only instead of several.
Is going colocation a natural progression of the web hosting business cycle? We start off with a Reseller Account in the beginning - then grow and lease a Dedicated Server. And then grow and lease another and another..... Is colocation the next big step?
I'm considering going with CalPop for one of my colo sites. I've read the user reviews and apart from those who were chaffed because of scratched servers and reboots they seem decent enough. My take on sticking anything in a datacenter is to stick it in yourself and rent the entire rack. That should take care of 90% of the negative reviews I have read, so my question is does any one have any real review of their services? Bandwidth performance, etc?
Secondly, I need a second colo for a redundant server. I'd like to find something closer on the East coast that offers the same pricing structure and services.
Anyone know what firewall do I need for my colo? I want to protect external IP. Here is my setup
3 servers, all have two nic cards, one of the cards will be the external IP and the other one will be LAN IP. So my question is what hardware firewall do I need to protect the External IP?
I was thinking of a cisco pix 515e. Which only route external IP to the LAN IP. I need something where I don't have to route, It just protect the external IP.
I need 1U of space, .5 amp (50 watts) power, one IP, 2.5mbps bidirectional bandwidth (total of 5mbps up + down) and about 10GB of traffic per day each direction (total of 20GB up + down). Would be nice if they have remote KVM along with console (serial) access. Location should be anywhere in USA.
Purpose is to host a VPN router for various remote locations to connect in to. Reliability and good connection (low latency) is important.
I've been a colo since the beginning of time. My servers are getting old so I've started pricing options, and it looks like dedicated is the way to go today. But I'm not sure...
I suppose it depends on the host. My host says "if you're colo, we provide admin at an hourly rate. If your machine needs a reboot, call us and we reboot it. If you're dedicated we don't touch your server beyond repairing it. If it needs a reboot, you login to our site and click a link and it is rebooted."
That doesn't seem like much of a difference. I'd need an off-site admin, but both charge by the hour, so no big deal. A live person reboot seems no better than a web-based software reboot. In fact I'll wager that the "live person" just logs in and clicks the link for me.
Colo is about twice as expensive as dedicated. That seems like the only big difference.
It also seems to me that with today's cPanel-style admin it's trivial to migrate to a new host, so competition to keep clients is intense. I'm guessing that keeps prices down. Reading between the lines of what my host says, I can tell he doesn't really want me to go dedicated. He kinda said they don't make much money on dedicated machines.
I have a web hosting business that has been growing constantly for a couple of years, now I think it's time to move on and instead of leasing servers start with colocation and operating my own hardware, I'm in Mexico and there are no good deals here as there are in other countries, so I have several questions about your appreciation of where things are better for business, US or Canada:
Where is colo cheaper? Where is hardware cheaper? Where is personnel less costly? Where is personnel better qualified? Where is office space cheaper? Where is electrical energy cheaper? Where are more investment opportunities? Where are taxes lower? Where is living less costly?
I've been researching these on my own but still have several doubts, may be you find some of this question obvious however I'd still like to know your appreciation.
Are there any established providers that offer this? I want an established company who isn't going to run off with my server, and they need to provide high quality bandwidth(Level3, AT&T, Sprint, etc) with 100% Power SLA, etc. in a secure facility.
IE: If I buy my own servers can I ship them out to a datacenter to have them fully managed, basically being a dedicated server? Except i'll own the hardware, and they provide the bandwidth, management, etc.
