OpticServers LTD goes live with owned Datacenter! Uk Based

[iWF-Jacob]

Yes i know we currently only have 2 servers under it and they are both limited to 1Gbps and each VPS is limited to 100Mbps this is a 1 Gbps Port and it is actually one of the higher end netgear models we will change it to a very high end router as soon as we can justify it for now it is under no load at all and handling the traffic very well but yes i fully understand that we do eventually need to change our core switch.

Seriously though, I'm all for you starting this and I think that's great but using that Netgear gs724t is not starting out on the right foot. In my other job in our "datacenter" which really is NOT a datacenter, just a room with a keypad, two UPS sets, two gen sets, and servers/networking equipment has a better core infrastructure then that, here's our 4500-X stack during a maintenance (hence not everything is plugged in). 

 

[OpticServers]

See, I genuinely don't think you do. Looking at the kit you have already in your single rack, none of that constitutes anything near being worthy of 'DC' status or even capable. If you're saying it works out better for you in terms of price then I can't see how you hope to expand to any degree when you're starting off with poor kit that 'fits' your 'price'. I see no large power feeds or commando sockets (which is the norm). No proper HVAC, just some fans and sound proofing.


I don't think you fully understand the concept of a 'datacenter' or indeed what's actually involved in setting one up and running it. The whole "we'll upgrade when it's needed" line is seriously cringe worthy.


I get the fact you're excited about doing this and the fact you have the premises is great. Passion is good when it comes to new ventures but ignorance isn't bliss. If you're not going to do it properly from the outset then you're only creating more work for yourselves in the long run and setting yourselves up for a very steep fall as well as letting any customers you somehow get down quite badly. I don't want to cause too much offence here but I genuinely believe you don't know what you're doing here.

Hey MartinD,

As i have said many times before we are starting this from our own pockets. so clearly we are not going to buy 15 full sized racks just because we want to make it look more like a datacenter and as i said before our networking gear is currently handling our network load fine once we need to upgrade hardware of course we will but we need to be able to justify putting a large amount of money on a switch or other networking gear. and the price part of it i did not mean just pricing of course this is not a cheaper option but in the long long in terms of comparing price to quality and control this was a much better option. and about the commando sockets, they are just 3 phase sockets.. what kind of server do you use that runs 3 phase ? we have a dedicated 100 AMP Single phase curcuit in our room and that is MORE then enough to supply 2+ full racks depending on how we share the load and there can be another 4 100 AMP Curcuits run into the room when we need them. and yes we do not have a "proper" HVAC but we have 2 intakes into the room that have large filters in between the walls of the intake vents that filters the air very well and then we have 2 top high powered fans that can do up to 10 full air changes an hour. and also the last part about setting up and making more work for our selfs? once we grow to a big enough client base we will be moving to our own warehouse and getting actual air conditioning units and doing it "proper" but for what we are doing right now we have more then enough capacity.

Nice site and panel, is that Virtkick?

No its fully custom coded

Seriously though, I'm all for you starting this and I think that's great but using that Netgear gs724t is not starting out on the right foot. In my other job in our "datacenter" which really is NOT a datacenter, just a room with a keypad, two UPS sets, two gen sets, and servers/networking equipment has a better core infrastructure then that, here's our 4500-X stack during a maintenance (hence not everything is plugged in). 

as i have said many times before this is just to start up with a couple of our servers once we grow were the switch is even getting close to 80% of its capacity we will change it over to a much larger capacity switch. but i do understand your point fully.

 

[MartinD]

Hey MartinD,

As i have said many times before we are starting this from our own pockets. so clearly we are not going to buy 15 full sized racks just because we want to make it look more like a datacenter and as i said before our networking gear is currently handling our network load fine once we need to upgrade hardware of course we will but we need to be able to justify putting a large amount of money on a switch or other networking gear.

I'll go back to the bit where I mentioned you not understanding or knowing what you're doing here. You cannot say "it's fine for now and we'll upgrade it later if needs be" when you're talking about your CORE NETWORK SWITCH. If it was some crap OOB switch or similar then sure, I can get with that but when you're talking about CORE NETWORK COMPONENTS you cannot be so blase. You will need to schedule a 100% outage for this - why is this so difficult to understand and why do you feel that it's acceptable to have a 100% outage just because you can't be bothered doing it properly from the outset? A few extra hundred quid and you'd have something a LOT more capable in place there that would last you a LOT longer than the desktop PoS you have just now.

It's also worth noting that when people search for OpticServers in future when deciding if you're suitable for their needs, threads like this will pop up.. and they will see that you thought a cheap-ass switch was perfectly fine as a CORE DC SWITCH because you paid for it "out of our own pockets" and that you can "upgrade when you need to" and you're perfectly happy for your lack of understanding to result in 100% outages for all your customers.

Edit: Also, just to add. A lot of us have passed similar comments about the switch. We can't all be wrong here. We are trying to give you the best advice and help you resolve a big problem before you get too deep. We've been there before, it's the collective voice of experience. I would urge you to listen and take action instead of throwing excuses out. We know what we're talking about - allow us to HELP you!

 

[splitice]

@iWF-Jacob That image doesnt load here, I was curious to see that.

One thing I missed when I was making my post, be careful about advertising DDoS protection as one of your key features. You will for sure attract those who get regular attacks, regular attacks that will likely cause instability due to either slow mitigation activation (network level filters in industry are often 30-90s for activation) or attacks that miss the filters and saturate a 2Gbps pipe. Sure its good to have, just be careful is what I am saying. I for one hope you continue to work on this area, you could be pretty viable when you reach a more realistic scale... and replace that core switch... I think everyone agrees regarding the core switch.

All in all I remain positive, just be careful.

 

[OpticServers]

I'll go back to the bit where I mentioned you not understanding or knowing what you're doing here. You cannot say "it's fine for now and we'll upgrade it later if needs be" when you're talking about your CORE NETWORK SWITCH. If it was some crap OOB switch or similar then sure, I can get with that but when you're talking about CORE NETWORK COMPONENTS you cannot be so blase. You will need to schedule a 100% outage for this - why is this so difficult to understand and why do you feel that it's acceptable to have a 100% outage just because you can't be bothered doing it properly from the outset? A few extra hundred quid and you'd have something a LOT more capable in place there that would last you a LOT longer than the desktop PoS you have just now.


It's also worth noting that when people search for OpticServers in future when deciding if you're suitable for their needs, threads like this will pop up.. and they will see that you thought a cheap-ass switch was perfectly fine as a CORE DC SWITCH because you paid for it "out of our own pockets" and that you can "upgrade when you need to" and you're perfectly happy for your lack of understanding to result in 100% outages for all your customers.


Edit: Also, just to add. A lot of us have passed similar comments about the switch. We can't all be wrong here. We are trying to give you the best advice and help you resolve a big problem before you get too deep. We've been there before, it's the collective voice of experience. I would urge you to listen and take action instead of throwing excuses out. We know what we're talking about - allow us to HELP you!

i never said it was not an issue to have a switch like this for a whole network on maybe 1 full rack or even two full racks but when i mean justify getting a better switch i mean as in once we get a nice amount of orders and we currently are meeting our targets we will start to upgrade hardware as ALL datacenters do datacenters constantly are switching our switch's and some even switch out core switch's yes we will have to schedule and alert everyone about the downtime but it will be around 30 - 60 seconds of outage of the network but again as i said before i do totally understand were you are coming from but for now this is capable for our network and will be for at least another couple of dedicated servers. and i clearly know what i am doing or i would not have gotten this way surely?

@iWF-Jacob That image doesnt load here, I was curious to see that.


One thing I missed when I was making my post, be careful about advertising DDoS protection as one of your key features. You will for sure attract those who get regular attacks, regular attacks that will likely cause instability due to either slow mitigation activation (network level filters in industry are often 30-90s for activation) or attacks that miss the filters and saturate a 2Gbps pipe. Sure its good to have, just be careful is what I am saying. I for one hope you continue to work on this area, you could be pretty viable when you reach a more realistic scale... and replace that core switch... I think everyone agrees regarding the core switch.


All in all I remain positive, just be careful.

Thank you for your advice and yea i understand but we have done some testing and there edge is actually constantly monitoring the network and normally starts to filter attacks with in 10 - 15 seconds longest we have seen is around 45 seconds and actually its good you brought this up because in around an hour or two we are going down to the datacenter to rack a dedicated server and i will be setting it up as a perimeter firewall to filter some of the more complex based attacks that some times get past the upstream's edge filtering system and auto a auto nullroute system so if a IP is going over a certain limit we will set and it is causing packet loss for other IP's it will get nullrouted temporarily and then every minute check if the attack has subsided and if not it will be re nullrouted so no other clients are effected by other clients IP's being flooded that exceeds our Limits

 

[iWF-Jacob]

@iWF-Jacob That image doesnt load here, I was curious to see that.

Strange, here you go: http://i.imgur.com/TP4daWx.jpg 

 

[Kenshin]

and i clearly know what i am doing or i would not have gotten this way surely?

If you really know what you're doing, you would have avoided showing off your datacentre or network uplinks as your advantages here, they aren't.

With the kind of setup you have, your target audience should be local small businesses that don't need high bandwidth but need either near physical access to the servers or managed support which you can provide since the facility is in-house for you. 2x 1G backbone isn't exactly huge to most people here who are used to having servers from OVH, Online, Leaseweb with Nx10G backbones. 2x 1G can be easily flooded with legit traffic and your "core switch" isn't going to be able to help you properly manage the bandwidth.

I stress, setting up your own datacentre and all is a good thing, there's definitely potential there, but the way you're trying to market yourself here is just making yourself a huge target for criticism. Even companies like Colocrossing were taken to task when photos of their network infrastructure was leaked. There are quite a number of people here who are network/server veterans who've worked on enterprise/carrier networks, you're not going to impress them with a setup like this.

And for the record, I was managing 10 racks in Singapore on just 300M capacity just 2 years ago spread across multiple datacentres (the agony), so I fully understand how proud you are of your network and having your own datacentre, but again, target audience vs marketing techniques.

 

[OpticServers]

If you really know what you're doing, you would have avoided showing off your datacentre or network uplinks as your advantages here, they aren't.

With the kind of setup you have, your target audience should be local small businesses that don't need high bandwidth but need either near physical access to the servers or managed support which you can provide since the facility is in-house for you. 2x 1G backbone isn't exactly huge to most people here who are used to having servers from OVH, Online, Leaseweb with Nx10G backbones. 2x 1G can be easily flooded with legit traffic and your "core switch" isn't going to be able to help you properly manage the bandwidth.

I stress, setting up your own datacentre and all is a good thing, there's definitely potential there, but the way you're trying to market yourself here is just making yourself a huge target for criticism. Even companies like Colocrossing were taken to task when photos of their network infrastructure was leaked. There are quite a number of people here who are network/server veterans who've worked on enterprise/carrier networks, you're not going to impress them with a setup like this.

And for the record, I was managing 10 racks in Singapore on just 300M capacity just 2 years ago spread across multiple datacentres (the agony), so I fully understand how proud you are of your network and having your own datacentre, but again, target audience vs marketing techniques.

Thank you for that.. i am defiantly taking on your advice. and the pictures and bandwidth capacity was not so much showing off more just showing people what we have and our 2 x 1Gbps Uplinks have 10Gbps Bearers so we can upgrade as soon as we need to.

 

[gordonrp]

Gotta start somewhere. Good luck. 

 

[gordonrp]

Also, is the building your dad's or something? If so, you're lucky to have your own facility. It will remove a ton of the BS that you may have to deal with regarding large corporate colos (scheduling power installs, arguing over cab to cab crossconnect rates, etc). 

I foresee a couple of issues with your plan;

• you will always be tied to virginmedia or whomever for IP transport (when you want to mix in other IP transit providers). Transporting SUCKS because it really adds to your costs, vs being in a well connected DC.
• No generator or UPS backup it seems

Dont fret, focus on selling based on CUSTOMER SERVICE, grow your business, and then upgrade the infra as cash flow allows. I agree with those above that your "core" sucks, but you should see if you can get any customers first. I would focus on small biz/webdesign type customers, or hosted email, etc, the low bandwidth type of clients, or local clients that you can sell to based on a handshake and a smile.

Good luck

 

[OpticServers]

Also, is the building your dad's or something? If so, you're lucky to have your own facility. It will remove a ton of the BS that you may have to deal with regarding large corporate colos (scheduling power installs, arguing over cab to cab crossconnect rates, etc). 

I foresee a couple of issues with your plan;

• you will always be tied to virginmedia or whomever for IP transport (when you want to mix in other IP transit providers). Transporting SUCKS because it really adds to your costs, vs being in a well connected DC.
• No generator or UPS backup it seems

Dont fret, focus on selling based on CUSTOMER SERVICE, grow your business, and then upgrade the infra as cash flow allows. I agree with those above that your "core" sucks, but you should see if you can get any customers first. I would focus on small biz/webdesign type customers, or hosted email, etc, the low bandwidth type of clients, or local clients that you can sell to based on a handshake and a smile.

Good luck

We do not own the building however Toby has a 8 year lease on this building so yes it is basically ours for the foreseeable future, We do have a UPS Backup system hooked into our main 300 AMP Curcuit on the building we will get some smaller backup UPS's for the room for extra backup time but we should get at least 4 - 6 hours of back up time under a full rack load depending on servers load's and other variables. and yes our core is not the best and can certainly be upgraded a lot and we defiantly will.. we are currently putting a edge filtering IDS System in play so we can filter some of the more complex Attacks that Virgin's Edge does not catch, but thank you for your advice and opinion i totally understand.

EDIT: we are also currently working out a long term deal with a hardware seller to get bulk hardware for selling Dedicated Servers and we are also working on offering Colocation!

 

[HalfEatenPie]

Also, is the building your dad's or something? If so, you're lucky to have your own facility. It will remove a ton of the BS that you may have to deal with regarding large corporate colos (scheduling power installs, arguing over cab to cab crossconnect rates, etc). 

I foresee a couple of issues with your plan;

• you will always be tied to virginmedia or whomever for IP transport (when you want to mix in other IP transit providers). Transporting SUCKS because it really adds to your costs, vs being in a well connected DC.
• No generator or UPS backup it seems

Dont fret, focus on selling based on CUSTOMER SERVICE, grow your business, and then upgrade the infra as cash flow allows. I agree with those above that your "core" sucks, but you should see if you can get any customers first. I would focus on small biz/webdesign type customers, or hosted email, etc, the low bandwidth type of clients, or local clients that you can sell to based on a handshake and a smile.

Good luck

I love this.

So spot on.

 

[RTGHM]

I was doing some digging, and you're a spawn of a hackforums host.

Sorry, but ads on hackforums lose my trust pretty quickly.

Just my two cents.

 

[DomainBop]

I was doing some digging, and you're a spawn of a hackforums host.

Sorry, but ads on hackforums lose my trust pretty quickly.

Just my two cents.

Does this mean you don't want a copy of Cheesebus 's "How to Infect Your Neighbor" ebook?

cheesebus:

(November 2012): "This ebook will learn you how to infect your neighbors around you"  <--offering an ebook

(October 2012) "I weared this signature for 1 week now and i made more then 50 posts I would like to get my free boot time please " <--responding to an offer for 2 free hours use of TruBoot booter

(April 2014) "Right now the best thing is to get your own Linux bots to attack. Booters on HF are barely 1 Gbps and some good ones are like 3-4 Gbps. PM me If you want some Linux bots." <--offering booter services

(March 2014) "1 am the owner of OpticServers."

OpticServers is currently being advertised and represented on HF by user HDPixel so I'm not sure if "Cheesebus" still owns it or is still involved in it, but either way I would never buy from a HF advertiser, especially one with such a sketchy history

 

[trewq]

Does this mean you don't want a copy of Cheesebus 's "How to Infect Your Neighbor" ebook?


cheesebus:


(November 2012): "This ebook will learn you how to infect your neighbors around you" <--offering an ebook

(October 2012) "I weared this signature for 1 week now and i made more then 50 posts I would like to get my free boot time please " <--responding to an offer for 2 free hours use of TruBoot booter


(April 2014) "Right now the best thing is to get your own Linux bots to attack. Booters on HF are barely 1 Gbps and some good ones are like 3-4 Gbps. PM me If you want some Linux bots." <--offering booter services


(March 2014) "1 am the owner of OpticServers."


OpticServers is currently being advertised and represented on HF by user HDPixel so I'm not sure if "Cheesebus" still owns it or is still involved in it, but either way I would never buy from a HF advertiser, especially one with such a sketchy history

I can't believe people on HF talk about booters as if they are just a normal thing. It's like yelling out the window you want to buy crack and dealers lining up at your door.

How these people have not been arrested yet I have no idea...

 

[RTGHM]

I can't believe people on HF talk about boosters as I'd they are just a normal thing. It's like yelling out the window you want to buy crack and dealers lining up at your door.


How these people have not been arrested yet I have no idea...

See: Brian Krebs Blackhat

See: NPT (Noob persistent threat)

Summary: Hackforums kids are not important to law enforcement. law enforcement goes after child peddlers, people who hack financial institutions, etc. not some folks who can pay $5 to hit off a website via paypal.

 

[OpticServers]

I was doing some digging, and you're a spawn of a hackforums host.

Sorry, but ads on hackforums lose my trust pretty quickly.

Just my two cents.

We are not a "hackforums host" it is just another advertising forum as we are concerned. you can see our Registered Company Documents here: http://data.companieshouse.gov.uk/doc/company/09200562

Does this mean you don't want a copy of Cheesebus 's "How to Infect Your Neighbor" ebook?

cheesebus:

(November 2012): "This ebook will learn you how to infect your neighbors around you"  <--offering an ebook

(October 2012) "I weared this signature for 1 week now and i made more then 50 posts I would like to get my free boot time please " <--responding to an offer for 2 free hours use of TruBoot booter

(April 2014) "Right now the best thing is to get your own Linux bots to attack. Booters on HF are barely 1 Gbps and some good ones are like 3-4 Gbps. PM me If you want some Linux bots." <--offering booter services

(March 2014) "1 am the owner of OpticServers."

OpticServers is currently being advertised and represented on HF by user HDPixel so I'm not sure if "Cheesebus" still owns it or is still involved in it, but either way I would never buy from a HF advertiser, especially one with such a sketchy history

i am not quite sure who cheesebus is. i own the account HDPixel and i have just under 300 + Reputation with no Negatives or neutrals.

EDIT: just to clarity the username on hackforums "HDPixel" IS me and i have never heard of chessebus nor is he affiliated with OpticServers LTD in anyway.

 

[RTGHM]

We are not a "hackforums host" it is just another advertising forum as we are concerned. you can see our Registered Company Documents here: http://data.companieshouse.gov.uk/doc/company/09200562

i am not quite sure who cheesebus is. i own the account HDPixel and i have just under 300 + Reputation with no Negatives or neutrals.

EDIT: just to clarity the username on hackforums "HDPixel" IS me and i have never heard of chessebus nor is he affiliated with OpticServers LTD in anyway.

Regardless if you think it's a "advertising forum" - I'm just saying a lot of people refuse to host with a company affiliated with hackforums whom is advertising on their website

Also, stating that you have 300+ reputation, really isn't helping you case. It just is making you look like an even bigger startup-and-fail operation like 99.9% of hackforums.

 

[OpticServers]

Regardless if you think it's a "advertising forum" - I'm just saying a lot of people refuse to host with a company affiliated with hackforums whom is advertising on their website

Also, stating that you have 300+ reputation, really isn't helping you case. It just is making you look like an even bigger startup-and-fail operation like 99.9% of hackforums.

that is understandable about the affiliation with HackForums, if it is really such a big issue then we may think about removing our thread from Hackforums. it was just another forums we could advertise on. 

 

[VPN.SH]

that is understandable about the affiliation with HackForums, if it is really such a big issue then we may think about removing our thread from Hackforums. it was just another forums we could advertise on. 

Just had a look at that HF thread. I imagine that a fair amount of people around here will avoid using your services if they know that you're advertising there. It generally presents a risk that people don't want to take on board, and also some people morally disagree with hosts posting there in the first place. Just a heads up .

Also, in that thread you mentioned moving your website to OpticServers, and somebody was complaining about it being hosted at OVH. It's fairly standard practice to host your website outside of your own infrastructure, as otherwise if you have issues with your DC, then a client is going to be wondering "Why is my server down? I'll go check out the OpticServers website" only to find out that your site is also down. If your billing/support panel is on the same system then clients won't even be able to shout up to find out what's happening.

Basically, it's less hassle to host your website and support outside of your own infrastructure.