I'm about to start a new django project with friends. This project includes a website, API to mobile applications, clearing and billing, internationalization aspects etc...
我即将与朋友们开始一个新的django项目。该项目包括一个网站,移动应用程序的API,清算和计费,国际化方面等......
What are the pros and cons for using Google App Engine vs. a regular web hosting solution?
使用Google App Engine与常规网络托管解决方案有什么优缺点?
Of course GAE documentation praises their support for django, but I would like to hear from people with first hand experience.
当然,GAE文件赞扬他们对django的支持,但我想听取有亲身经验的人的意见。
Here's a general discussion regarding pros\cons for GAE. I'm more interested in the details regarding django (and related libraries) support.
以下是关于GAE的优缺点的一般性讨论。我对django(和相关库)支持的细节更感兴趣。
2 个解决方案
#1
8
What sold me on Google App Engine vs hosting a Django app were the following points:
在Google App Engine上销售我而不是托管Django应用程序的原因有以下几点:
- No DB Admin: We didn't have to worry about administering and managing a database. This was super productive for a two man team.
- The Datastore Makes Sense: Given that we had no RDBMS experience, the GAE Datastore just made a lot of sense: It is a big hash table with transaction support. Definitely less complex -usage wise- than a traditional database.
- No Sysadmin Skills Needed: Not sure about you, but we did not have mad sysadmin skills. We just wrote a simple Python script that compressed our JavaScript and CSS and deployed. That was good enough.
- Easy Deployment and Versioning: The online interface allows you to test out a new version before making it the default. So if the new version breaks, you can easily roll to an older version.
- Easy Scalability: We enabled billing for up to $4 a day, which gave us a lot of CPU cycles and a huge email quota. We didn't have to worry about servers going down. Again, this give you peace of mind if you're a small team.
没有数据库管理员:我们不必担心管理和管理数据库。对于一个双人球队来说,这是超级高效的。
数据存储区有意义:鉴于我们没有RDBMS经验,GAE数据存储区很有意义:它是一个具有事务支持的大型哈希表。绝对不如传统数据库那么复杂。
不需要系统管理员技能:不确定你,但我们没有疯狂的系统管理员技能。我们刚刚写了一个简单的Python脚本来压缩我们的JavaScript和CSS并进行部署。那太好了。
轻松部署和版本控制:在线界面允许您在将其作为默认版本之前测试新版本。因此,如果新版本中断,您可以轻松滚动到旧版本。
易于扩展:我们每天最多支付4美元的计费,这给了我们很多CPU周期和巨大的电子邮件配额。我们不必担心服务器出现故障。如果你是一个小团队,这可以让你高枕无忧。
You also get easy access to cron, offline tasks and email. Oh, and you can also use Django's templating engine, which is one of my favorite bits of Django.
您还可以轻松访问cron,离线任务和电子邮件。哦,你也可以使用Django的模板引擎,这是我最喜欢的Django之一。
Cons:
- No FTP support: A lot of enterprises still use FTP. You can only make HTTP/S requrests with GAE, so you can't serve or upload FTP files.
- 3000 files per app: Your app cannot have more than that. But for the Python version, you can bundle up extra libraries in zip files and use zipserve to serve them. I usually also sprite up small images into one and compress JavaScript and CSS and lump them into one file each.
- No expensive calculations: All requests have to finish up in 30 seconds. But GAE gives you Cron and task queues, so if you need to do super expensive calculations, then you'll have to break them up.
- Only pure Python libraries: So you can't use cPickle for example.
- No legit image processing: They image processing API does provide a subset of PIL, but if you need to do heavy lifting you'd better take it outside (of GAE).
没有FTP支持:很多企业仍然使用FTP。您只能使用GAE进行HTTP / S请求,因此您无法提供或上传FTP文件。
每个应用3000个文件:您的应用不能超过这个数量。但是对于Python版本,您可以在zip文件中捆绑额外的库,并使用zipserve来提供它们。我通常也会将小图像编成一个并压缩JavaScript和CSS,并将它们分别归为一个文件。
没有昂贵的计算:所有请求都必须在30秒内完成。但GAE为您提供了Cron和任务队列,因此如果您需要进行超级昂贵的计算,那么您将不得不将它们分解。
只有纯Python库:所以你不能使用cPickle作为例子。
没有合法的图像处理:他们的图像处理API确实提供了PIL的子集,但是如果你需要做大量的提升,你最好把它带到外面(GAE)。
I would highly recommend GAE if you have a small team - which sounds like you do, and you do not like to mess around with configuring servers. .
如果你有一个小团队,我强烈推荐GAE - 这听起来像你一样,你不喜欢乱搞配置服务器。 。
Best of luck!
祝你好运!
#2
2
Its free quotas are quite generous, so you wouldn't be paying anything until your website starts receiving a considerable amount of visits. Which makes it ideal for a project like yours, where you might not know whether it's going to be a success.
它的免费配额非常慷慨,所以在您的网站开始接受大量访问之前,您不会支付任何费用。这使得它非常适合像您这样的项目,在那里您可能不知道它是否会取得成功。
#1
8
What sold me on Google App Engine vs hosting a Django app were the following points:
在Google App Engine上销售我而不是托管Django应用程序的原因有以下几点:
- No DB Admin: We didn't have to worry about administering and managing a database. This was super productive for a two man team.
- The Datastore Makes Sense: Given that we had no RDBMS experience, the GAE Datastore just made a lot of sense: It is a big hash table with transaction support. Definitely less complex -usage wise- than a traditional database.
- No Sysadmin Skills Needed: Not sure about you, but we did not have mad sysadmin skills. We just wrote a simple Python script that compressed our JavaScript and CSS and deployed. That was good enough.
- Easy Deployment and Versioning: The online interface allows you to test out a new version before making it the default. So if the new version breaks, you can easily roll to an older version.
- Easy Scalability: We enabled billing for up to $4 a day, which gave us a lot of CPU cycles and a huge email quota. We didn't have to worry about servers going down. Again, this give you peace of mind if you're a small team.
没有数据库管理员:我们不必担心管理和管理数据库。对于一个双人球队来说,这是超级高效的。
数据存储区有意义:鉴于我们没有RDBMS经验,GAE数据存储区很有意义:它是一个具有事务支持的大型哈希表。绝对不如传统数据库那么复杂。
不需要系统管理员技能:不确定你,但我们没有疯狂的系统管理员技能。我们刚刚写了一个简单的Python脚本来压缩我们的JavaScript和CSS并进行部署。那太好了。
轻松部署和版本控制:在线界面允许您在将其作为默认版本之前测试新版本。因此,如果新版本中断,您可以轻松滚动到旧版本。
易于扩展:我们每天最多支付4美元的计费,这给了我们很多CPU周期和巨大的电子邮件配额。我们不必担心服务器出现故障。如果你是一个小团队,这可以让你高枕无忧。
You also get easy access to cron, offline tasks and email. Oh, and you can also use Django's templating engine, which is one of my favorite bits of Django.
您还可以轻松访问cron,离线任务和电子邮件。哦,你也可以使用Django的模板引擎,这是我最喜欢的Django之一。
Cons:
- No FTP support: A lot of enterprises still use FTP. You can only make HTTP/S requrests with GAE, so you can't serve or upload FTP files.
- 3000 files per app: Your app cannot have more than that. But for the Python version, you can bundle up extra libraries in zip files and use zipserve to serve them. I usually also sprite up small images into one and compress JavaScript and CSS and lump them into one file each.
- No expensive calculations: All requests have to finish up in 30 seconds. But GAE gives you Cron and task queues, so if you need to do super expensive calculations, then you'll have to break them up.
- Only pure Python libraries: So you can't use cPickle for example.
- No legit image processing: They image processing API does provide a subset of PIL, but if you need to do heavy lifting you'd better take it outside (of GAE).
没有FTP支持:很多企业仍然使用FTP。您只能使用GAE进行HTTP / S请求,因此您无法提供或上传FTP文件。
每个应用3000个文件:您的应用不能超过这个数量。但是对于Python版本,您可以在zip文件中捆绑额外的库,并使用zipserve来提供它们。我通常也会将小图像编成一个并压缩JavaScript和CSS,并将它们分别归为一个文件。
没有昂贵的计算:所有请求都必须在30秒内完成。但GAE为您提供了Cron和任务队列,因此如果您需要进行超级昂贵的计算,那么您将不得不将它们分解。
只有纯Python库:所以你不能使用cPickle作为例子。
没有合法的图像处理:他们的图像处理API确实提供了PIL的子集,但是如果你需要做大量的提升,你最好把它带到外面(GAE)。
I would highly recommend GAE if you have a small team - which sounds like you do, and you do not like to mess around with configuring servers. .
如果你有一个小团队,我强烈推荐GAE - 这听起来像你一样,你不喜欢乱搞配置服务器。 。
Best of luck!
祝你好运!
#2
2
Its free quotas are quite generous, so you wouldn't be paying anything until your website starts receiving a considerable amount of visits. Which makes it ideal for a project like yours, where you might not know whether it's going to be a success.
它的免费配额非常慷慨,所以在您的网站开始接受大量访问之前,您不会支付任何费用。这使得它非常适合像您这样的项目,在那里您可能不知道它是否会取得成功。