What is the minimum and maximum CPU value for a single server?
The minimum CPU value for a single server is 500 core-MHz and the maximum CPU value is 20000 core-MHz.
What do you mean by core-MHz?
The effective speed of modern CPUs is determined both by the clock speed of the processor (in MHz) and the number of independent execution units (cores). For a given processor design, the available CPU bandwidth is roughly proportional to the product of these two. Concretely, if you create a virtual server with 2100 core-MHz of CPU allocated, you should expect performance roughly equivalent to a single core of AMD Opteron 2352 clocked at 2.1GHz, and this performance varies linearly in proportion to the core-MHz number.
How do I set the number of CPU cores?
This is under the advanced settings on the server configuration page – simlpy click Edit when your server is powered off to view and edit these settings.
What is the minimum and maximum memory for a single server?
The minimum memory value for a single server is 256 MB RAM.
All our zones support servers running 32GB RAM, except our Sat-P zone, where the limit is 8GB.
What is the minimum and maximum size of a single drive?
The minimum size is 1GB and the maximum size is 2 TB for a single drive.
Can I attach one drive to more than one server?
No. However, you can run filesharing protocols such as NFS or Windows File Sharing from one server to another over a private network VLAN.
How do I attach more than one drive to a server?
You can set this in the server configuration page. (Click Edit when your server is powered off to find this page).
What storage performance can I expect?
You should be able to achieve around 30-40MB/s streaming read and write performance from your virtual drives. However, contention with other users accessing the same underlying RAID arrays can result in lower performance, especially at peak times. This affects writes more than reads as we cannot safely cache disk writes as aggressively.
What is the difference between IDE, SCSI and VirtIO drives? Which do you recommend?
If you’re not sure then you should stick with the IDE emulation, which offers good performance and compatibility. However, we believe that VirtIO drive types provide the best performance, and these are what our preinstalled images use by default.
Note that our virtio drives have writeback caching enabled, so you will need Linux 2.6.32 or later, which supports barriers over VirtIO, to be able to use these safely. If you’re not using one of our recent preinstalled images, VirtIO disk drivers must be installed specifically within Windows before you can use this disk type. See our guide for switching drives to Virtio block devices.
What is dedicated disk storage, and what are the advantages?
Our normal virtual drives are allocated from RAID1 disk pairs shared between customers. This means that when one customer is heavily accessing their disks, your accesses may be slower if they happen to be on the same disk pair. If you need uncontended access to your virtual drives, we can allocate a RAID1 disk pair specifically for your use, with no other customer drives on it. This means that performance on your drives cannot be degraded by the activity of other ElasticHosts users.
What happens to my data when my server shuts down?
Data stored on your virtual drives is persistent and will be available when you restart your server, just as it would be on a physical hard disk. This means that you can install software and save data exactly as you would on a physical server, unlike some other cloud solutions.
How do I resize my server?
When your server is powered off, go into the server configuration page, then simply adjust the CPU and memory settings.
How do you recommend configuring server clusters appropriate for each stage in my website’s growth?
Whilst the exact answer will vary according to your web application’s architecture, we recommend the following cluster architectures to back a typical site as it grows, progressing in order:
- Single small server with standard disk, running entire LAMP stack
- Single large server with dedicated disk, running entire LAMP stack
- Multiple large servers in classic 3-tier web application architecture, interconnected over private network VLAN:
- Clustered database servers with dedicated disk
- Front-end application servers with standard disk
- Load balancer(s) running Pound, HA-proxy, etc.
Note that in the second stage a single larger server with dedicated disk will typically give better performance than splitting immediately into multiple small servers, since disk I/O is improved and there are no extra latencies from cross-network communication.
How do I scale automatically with load?
You can use our API to start and stop servers when your application detects that it needs extra computing power. Building an application which scales across multiple machines like this is done in the same way as it would be on physical hardware which you could programmatically turn on and off, but is beyond the scope of this FAQ.
How do I shrink my disk safely?
It is very important to shrink the size of your filesystem and any partitions using your operating system BEFORE trying to shrink the drive itself! If you don’t do this, you will chop the end off the filesystem when you shrink the drive leading to serious corruption.
If in doubt, we strongly advise against attempting a drive shrink because of this risk: this functionality is disabled in the web control panel to prevent accidental loss of data.
However, once you are sure it is safe to do so, you can shrink the virtual drive by shutting down the attached server and either decreasing the drive size via the API, or contacting the support team to shrink the drive as required.
How do I grow my disk?
To increase the size of your drive, shut down the server it is attached to, choose Edit for that drive in your control panel and increase the size in the dialog. After resizing the disk, you can restart your server attached to it but will then need to grow any partitions and filesystems on it to make use of the extra space. (Our pre-installed Linux images do not have partitions but just a single filesystem filling the entire drive, so it is sufficient to run resize2fs directly on the hard disk block device node to grow this filesystem).
Can I change the size of my server whilst it is running?
No. Operating systems are not able to recognise changes of this sort and would crash if we enabled them.