Tag Info

Hot answers tagged

13

There is nothing in a server that couldn't be in desktop hardware (assuming Intel servers.) Some things that are in a server that make it better to provide services 24x7 than desktop hardware: RAID and multiple disks, so that a single disk failure won't take down the server or lose data. Better disks, that can handle more and longer duty Better disks, that ...


10

If you NEED a big box like that I can recommend from first-hand-experience HP's DL580 G5/DL585 G6 24-core servers and the HP DL785 G6 48-core server. Please consider that buying any of these (or Dell/IBM/etc equivilant) machines RIGHT NOW is a very bad idea - the reason is that late this year or just into next year we will see new servers from all of these ...


9

Please help Dude, a reality check. Machines BREAK DOWN. When they break down, they get fixed or replaces. YOu must be VERY new to life if you never needed for example a car mechanic. Things DO break down, there is nothing special about it. Now, what to do? SIMPLE. Depending on when the server was purchased, do one of the following: contact shop for ...


8

The idea with ZFS is to let it known as much as possible how is disks are behaving. Then, from worst to better: Hardware raid (ZFS has absolutely no clue about the real hardware), JBOD mode (The issue being more about any potential expander: less bandwidth), HBA mode being the ideal (ZFS know everything about the disks) As ZFS is quite paranoid about ...


8

Ask them. Really. It can have subtle differences in meaning since people who don't know may coin the term or steal it and twist it for their own purposes. Usually it just means you installed and configured a system to act as a server. As them, clarifying that you'd build many Linux and Windows servers that acted as XYZ. There's nothing wrong in an ...


8

Although I'm sure you could find quieter fans I would be tempted not to bother messing with them. Dell have lots of engineers who spend their whole lives balancing the various physical requirements and constraints required to adequately cool the parts as laid out in their servers - messing with this cooling will almost certainly affect some of the components ...


7

Neither AD nor DNS are particularly CPU heavy and for the number of clients you're talking about you could easily get away with a dual-core chip, even at a lower speed than you're suggesting. MSQL will use all the cores you can throw at it pretty well, go for the quad core - oh and consider getting a dual-CPU E55xx series server for your DB, even if you ...


7

The Dell suite is called "Open Manage". You have a lot of info here on their site. Basically on a rpm-based distro ( in my experience this works very well with rhel-based distros) you just have to do: wget -q -O - http://linux.dell.com/repo/hardware/latest/bootstrap.cgi | bash yum install srvadmin-storage ( actually dell recommends srvadmin-all, but if ...


7

I would try the following command which might you give you the information you need depending on the motherboard: sudo dmidecode -t memory For example on one of my Dell servers I get the part number (but no so lucky on my workstation): Memory Device Array Handle: 0x1000 Error Information Handle: Not Provided Total Width: 72 bits ...


6

Your support company are dead on right - it's all fine now, but the day that it starts randomly blue screening will be the start of a mass panic, with no one to turn to. Do the upgrade now, while you have time to plan it and implement it properly - not when you've just found out that you can't buy a replacement motherboard. Supportability is important, ...


6

A great reason to change it is if you want to add another task to your list of things to do while increasing the chances of something going wrong. All joking aside, there really isn't any reason I've heard of to change the drive ahead of time. If you have RAID in place, you already have protection in place (assuming you have decent backups), and you're not ...


6

""What is it with server hardware that actually makes it more suited for server hosting vs. say a collection of cheap PCs or a good PC?"" When hosting a lot of servers, holding them up, getting power to them, being able to get to them all makes a difference. The typical way to do that is in a rack ( http://images.google.com/images?q=server%20rack ) and ...


5

A server is nothing more than a typical computer built to perform better under specific conditions (utilization, environment, performance, reliability, etc) than a typical computer. Typically they have specialized hardware to help accomplish this. A server rack is a container; it has rails on the four vertical corners to which equipment can be attached. ...


5

For the AD/DNS server for that few users, either will more than do. For the SQL server, then I would go for the quad core [unless it will only be running single queries most of the time (i.e. not large multi-user systems) in which case the higher speed of each core might be more beneficial than the extra cores] all other things being equal. "All other ...


5

I do hope you have more than a bare blade server. It's completely useless without the accompanying chassis and interconnects. The full specifications for the server are available here. These are the only hard drive options that will work for your server. They are all old-school 3.5" parallel SCSI disks. The BL20p G3 server can only accept two hard disks. ...


5

Q. If one happens to have some server-grade hardware at ones disposal, is it ever advisable to run ZFS on top of a hardware-based RAID1 or some such? A. It is strongly preferable to run ZFS straight to disk, and not make use of any form of RAID in between. Whether or not a system that effectively requires you make use of the RAID card precludes the use of ...


5

Yes, that's all correct. I suppose it's not clear from the Supermicro product pages. SAS SFF-8087 is a 4-lane multi-connection transport. It's typically used by RAID controllers and disk backplanes. You probably want to avoid anything that requires fanout cables. Expander backplanes can have some problems, depending on what types of disks and controllers ...


5

You really want to hire a consultant to work with you on this. If you do it wrong, your launch can sink your business after just a couple customers (and definitely after your first "incident"...any plans in place for backups? Link redundancy? Even security among your hosted sites and your management network?) Best advice for you; get a consultant, and start ...


5

If I were you, especially when starting out, I would look at leasing a virtual server that you can pay more or less for based on your resource usage & bandwidth usage. Once you have some real world statistics to use for trending you can use them to decide if/when moving away from cloud hosting is a sensible thing to do.


5

Neither. If virtual machine performance is paramount, you may want to consider RAID 1+0 with a hot-spare instead of one of the parity RAID variants. VM activity tends to be mixed random read/write activity. RAID 6 is poor at the random write game. RAID 5 should probably only be considered if you're space-constrained. RAID 10 Good when: You want speed and ...


5

You're going to buy a lot of servers that have to be "qualified" for an application that isn't written yet. That's like buying a truck for a purpose you don't know yet. There's really no way to "qualify" it unless you know what the requirements are. Usually qualified systems are those tested to work with a given application by using a tight set of spec'ed ...


5

That is correct. The card's interface is an SFF-8484. Depending on what your drive cage looks like, the other side of the cable needs to match. If you are doing this without hot-swap drives, you'll need to use 4-lane SAS breakout cables (e.g. SAS SFF-8484 to SFF-8482).


4

These are usually pre-release systems or components. I've seen them in the financial industry and used for benchmarking well before the product was available for consumers. I recall this specifically with the Nehalem and Nehalem-EX processor releases. These engineering sample items will work, and are stable. The main issue is that any faults or issues found ...


4

Easiest way for you get estimate and probably stay away from data center business is to talk to a co-location/hosting provider. They would give you estimate how much they would charge for a single rack, bandwidth, power etc.. You can also get quotes for 10 servers from some vendors and get your numbers together from these estimates.


4

Configuring RSA/IMT board would allow you to monitor Hardware events. Configuring RSA II: Configure RSA Board *PDF IBM Director can as well be configured to forward events even without the usage of IBM Director Manager, but this requires Tivoli Monitoring to capture Director events.


4

Right. Now we've got lots more information. You're running 3 sites, on 2 cores at 1.7GHz each. You've got 4GB of RAM, and about 3GB free (that's good). MySQL and Apache on the same server. About 20 httpd processes. A bunch of other stuff like nginx and named running. Is this server deliberately the jack of all trades? I think you need to scale out a ...


4

Backblaze has a business model that allows them to do this. They need gobs of storage space (though it doesn't have to be particularly fast) and they need it cheaply. It's worth it for them to dote on storage because they need so much and they can scale their production to handle having sufficient spare parts and support labor on-site. If your company's ...


4

lots of redundancy IMO Consumer grade hardware is not inherently evil, but you should certainly understand what you are getting and plan properly for their failure and replacement. There is a big difference between the really low-end consumer gear, and the high-end server gear, there is also some stuff in the middle that could be perfectly fine in some ...


4

pciconf -lv will give a list of the PCI devices in the system. It's not a great way to read chipset information and may return very little usable information, at any rate it will be in cryptic form. The dmidecode port would be much preferred, but pciconf is part of the base OS.



Only top voted, non community-wiki answers of a minimum length are eligible