|
 
- 帖子
- 102
- 积分
- 796
- 金钱
- 102
- 在线时间
- 112 小时
- 注册时间
- 2012-9-21
- 最后登录
- 2016-3-1
|
这么便宜啊
------------------------------------------------借楼
----12.13 收到一信，应证了我说的 自动重启。
他家只要负载过高，就会自动重启VPS 这样就能保证服务器稳定。
他以前是 负载超过2，时间超过10分钟就会自动重启。
现在修改为 负载超过2，时间超过1小时才会重启。
邮件中还说明了，自动重启可能会导致数据库出问题，所以重启时间延长至1小时。
Happy holidays! Crissic Solutions wants to say thank you for being a valued customer. This is a year-end email, so it will be rather lengthy.
The Idea Reborn
In April of this year, we restarted Crissic Solutions with a simple goal in mind – offer incredibly competitive services, but at a sustainable price. There are many insane offers out there - $30/yr KVM or $15/yr OpenVZ deals on rented hardware where they barely break even. Other companies pump yearly promotional sales regularly just to cover their hosting bills. Many of these companies die or vanish from the marketplace because they operate on a bad business model that is not sustainable or even profitable. We’ve managed to match and beat some of these “crazy” offers, but at a level that is both profitable and sustainable. We've done so without sacrificing performance and without sacrificing our high level of customer support.
Crissic Solutions Facts
December 26th, is our company 8-month anniversary
Clients currently in 85 countries
2,000 active services deployed
16 Dell C1100 servers online (and yes, we own them all)
15,360 IP addresses in our own ASN (http://bgp.he.net/AS62639)
250TB of bandwidth consumed monthly and growing
Cash flow positive and profitable
New Server Upgrades and Vrrrooom! Faster Disk IO
We’ve received lots of feedback about our services. Some comments made it clear that we had room for improvement. Those comments have been almost entirely related to our hard disk performance, or in many cases the lack thereof.
As of last week, we have stopped ordering C1100’s in favor of newer and more powerful servers. All new services will be provisioned on our latest hardware:
Dual Intel X5660 Hex Core 2.8Ghz CPUs
12 (yes TWELVE) Western Digital RE4 hard drives in RAID-10 with hardware RAID via a Dell H700 RAID card with battery backup
(Benchmarks show upwards of 600M/s disk performance, and significant load-time reduction due to no more IO wait!)
Upcoming Customer Server Upgrades
Beginning later next week, we will start a migration process, upgrading some (but not all – at least not right away) clients to our new hardware. The plan is to have all 2,000 services migrated to our new hardware in the next 3 to 6 months. A more detailed email with migration information will be provided as necessary. The first server to be migrated will be JAX5. We will send out an email sometime this weekend detailing a timeline. We will be performing live migrations so you should experience virtually zero service interruption.
Service Monitoring and Abuse System Modifications
We use a program called VPSMON to enforce specific abuse prevention rules. These rules limit SMTP connections to prevent IP blacklists, CPU restrictions to prevent node overload due to one or two customers, and load based restarts to keep the server operating smoothly. As a result of poor disk performance on existing nodes, services that had a decent amount of disk IO activity ended up getting rebooted as a result of these rules and disk IO wait.
These restart restrictions had to be put in place to make sure that all services on the node perform optimally. The unfortunate consequence is any service that utilizes a large chunk of IO (MySQL, active websites, etc.) gets restarted too frequently, and as a result negatively impacts a handful of users. Going forward, this restriction will still EXIST, but it will be a much lighter restriction (we’re looking at a load of 2.0 for about an hour or so to issue a reboot instead of 10 minutes) and it will be harder to hit as a result of IO performance gains. This means you will be able to use your service how you NEED and WANT to, rather than being restricted by rules that previously were hard to work within.
This restriction removal will NOT be in effect until all services are migrated. Based on existing services already migrated to new node, most users won’t hit any of the restart limitations. If you do end up hitting these limitations, we can set manual container configurations per-VPS (upon request) until the entire system rolls over to the new upgraded hardware.
This is just step one of many in the coming months. Our priority is improving our service quality without requiring dreadful price increases.
2014 New Year's Resolutions
Here are some of what we have planned to make Crissic even better during 2014, our New Year's Resolutions!
Re-focusing our attention on our KVM line with a server upgrades, new packages, and affordable price points just like we do with our OpenVZ line.
Enhancing the security of our billing and VPS portals
Better integration between billing and our panels
Incremental backups on a 24-hour time frame at no additional cost, and with the option to restore individual files, folders, or your entire VPS with a simple button click.
Shared and reseller clients, new nodes with CloudLinux and a plethora of additional addons
Thank you!
Without your support, we wouldn’t be where we are today. Thank you for choosing Crissic, believing in Crissic, and giving us the opportunity to serve as your VPS and Shared hosting provider.
As always, we value your feedback. If you have any feedback, good or bad, please do not hesitate to open a support ticket, or hop into IRC and let us know what you think!
Have a safe and happy holiday!
Crissic Solutions LLC
----------------------------GG翻译其中的
重生理念
今年四月，我们重新开始考虑一个简单的目标Crissic解决方案 - 提供令人难以置信的竞争力的服务，但在可持续发展的价格。有许多疯狂的报价在那里 - $ 30/yr KVM或$ 15/yr OpenVZ的交易在租用的硬件，他们勉强收支平衡。其他泵公司每年销售的促销经常只是为了掩饰他们的托管费。许多这些公司死亡或消失从市场，因为他们在一个糟糕的商业模式是不可持续的，甚至盈利的操作。我们已经成功地匹配，并击败其中的一些“疯狂”的计划书，但在一定的水平，既有利可图的和可持续的。我们已经在不牺牲性能和不牺牲我们的高水平的客户支持水平这样做。
Crissic解决方案的事实
12月26日，是我们公司8个月的纪念日
客户目前在85个国家
部署2000活动服务
16戴尔C1100服务器在线（是的，我们拥有他们全部）
在我们自己的ASN 15,360 IP地址（ http://bgp.he.net/AS62639 ）
带宽250TB每月消费和成长
正现金流和盈利能力
新的服务器升级和Vrrrooom ！更快的磁盘IO
我们已经收到了许多关于我们的服务的反馈。有些意见明确指出，我们有提升的空间。这些意见已几乎完全与我们的硬盘性能，或在许多情况下其缺乏。
截至上周，我们已经停止订购C1100的赞成更新，更强大的服务器。所有新的服务将置备于本公司最新的硬件：
双Intel X5660六角核2.8GHz的CPU
12 （是十二）西部数据RE4硬盘通过戴尔H700 RAID卡与电池备份驱动器的RAID-10与硬件RAID
（基准测试表明向上的600M / s的磁盘性能，并显著加载时间减少，由于没有更多的IO等待！ ）
即将到来的客户服务器升级
后来下周开始，我们将开始一个迁移过程，提升了一些（但不是全部 - 至少不是马上）客户对我们的新硬件。该计划是让所有2,000服务迁移到新的硬件，在未来3至6个月。与迁移信息更详细的电子邮件将被作为必要的。要迁移的第一个服务器将JAX5 。我们将发送一封电子邮件的某个周末，详细的时间表。我们将要执行实时迁移，所以你应该体验几乎是零服务中断。
服务监视和滥用系统修改
我们使用一个名为VPSMON程序来执行特定的滥用预防规则。这些规则限制的SMTP连接，以防止IP黑名单， CPU的限制，以防止过载节点由于一个或两个客户，并且根据负载来重新使服务器运行顺畅。由于对现有节点，服务，磁盘有一个像样的量的磁盘性能不佳导致IO活动最终获得重新启动作为这些规则和磁盘IO等待结果。
这些重新启动限制必须到位，以确保该节点上的所有服务达到最佳性能。不幸的后果是，利用IO （ MySQL的，积极的网站等）的一大块被重新启动过于频繁的任何服务，并因此产生了负面影响少数用户。展望未来，这种限制会依然存在，但是这将是一个轻得多的限制（我们正在寻找的2.0大约一个小时的负载或使发行重启，而不是10分钟） ，这将是很难打的IO的性能提升的结果。这意味着你将能够如何你需要和想要使用你的服务，而不是被限制，以前是很难在工作规则。
这个限制解除将不会生效，直到所有的服务都将迁移。根据已经迁移到新节点的现有服务，大多数用户不会碰到任何的重启局限性。如果最终打这些限制，我们可以（根据要求）设置每个VPS手动容器的配置，直到整个系统翻转到新升级的硬件。
这只是许多步骤，在未来的月份之一。我们的首要任务是提高我们的服务质量，而无需可怕的涨价。
2014新年的决议
下面是一些我们已经计划让Crissic更好的在2014年，我们的新年新希望！
重新集中我们的注意力在我们与一个服务器升级，新的软件包，并且就像我们做我们的OpenVZ的行合理的价位KVM线。
提高我们的帐单和VPS门户网站的安全性
计费和我们的面板之间更好的集成
增量备份在24小时的时间框架在不增加成本，并与恢复单个文件，文件夹或整个VPS用一个简单的按钮，点击该选项。
共享和经销商客户，与CloudLinux的和过多的附加插件的新节点 |
|
