Hosting/Webhosting MEGATHREAD

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Hosting/Webhosting MEGATHREAD

Fangs404: Dec 20, 2004; I time bomb.

Signed up for Linode and migrated my MediaTemple site there today. I haven't done much web administration stuff under Linux before, but it was really easy and straightforward thanks to Linode's amazing documentation. I just got the base 512mb plan, and it's much faster than MT was. The flexibility of a VPS is awesome.

I have encountered one problem, though. I have a WordPress blog installed. When I try to upload a plugin through the web interface ([domain]/wordpress/wp-admin/plugin-install.php?tab=upload), it doesn't work. I get taken to a page that wants my FTP login info. I assumed this is a permissions issue, so I chmodded 777 wp-content and all directories under that directory (plugins, themes, upgrade, and uploads). It still fails, though. I Googled some, and some people suggest to chown with the httpd user (www-data). I tried that just to see if it worked, and it did. I don't want to do that, though, because it's a pain in the rear end to upload files through SFTP, and it's also insecure. Does anyone know how to fix this without chowning all the files for the httpd process?

Fangs404 fucked around with this message at 11:38 on Mar 6, 2011

# ¿ Mar 6, 2011 11:21

Adbot: ADBOT LOVES YOU

# ¿ Apr 28, 2024 12:56

Fangs404: Dec 20, 2004; I time bomb.

Biowarfare posted:

get suphp working or something, iirc

Hey, this might be what I need. Thanks.

Bob Morales posted:

Don't 'chmod 777' and the whole reason you want to use sftp over ftp is because it's secure.

I think you misunderstood my problem.

eightysixed posted:

If you enter your FTP info, it will still install the plugin tho.

I don't actually have an FTP server running, and if possible, I'd like to keep it that way. I do all my file transfers over SSH (SFTP).

# ¿ Mar 6, 2011 20:34

Fangs404: Dec 20, 2004; I time bomb.

dvgrhl posted:

If that ends up being the case, post back if you would. I have the same issue with a website at work, and I just haven't had the time yet to look into what is needed to get automatic updates working.

It is indeed working. Documentation on how to setup suphp sucks, but http://www.pc-freak.net/blog/installing-suphp-on-debian-lenny-5-04-with-apache-2-2-9-2/ helped a lot. If you get an internal server error, know that you need to play around with the docroot and check_vhost_docroot settings in suphp.conf.

[edit]
I read that suphp runs about 25% slower than mod_php, so I decided to try to find a better solution without using suphp. Here's what I found:

WordPress has some suggestions for permissions. The important part is this:

quote:

All files should be owned by your user account on your web server, and should be writable by your username. Files should never be owned by the webserver process itself (sometimes this is www, or apache, or nobody).

Any file that needs write access from WordPress should be group-owned by the user account used by the webserver. For example, you may have a user account that lets you FTP files back and forth to your server, but your server itself may run using a separate user, in a separate usergroup, such as dhapache or nobody.

This alone, at least for me, wasn't quite enough to do the trick. What I discovered is that apparently WordPress checks to see if you are the user trying to write the files. Because the web user (in my case, www-data) is not the same as the owner of the files, it fails. It fails even if permissions are 777. The way to get beyond this is to add this line to your wp-config.php:

php:

<?
define('FS_METHOD', 'direct');?>

This indeed did the trick. So basically, add that line to your wp-config.php, and 644 all files and 755 all folders in your wordpress directory:

code:

find wordpress/ -type f -exec chmod 644 {} \;
find wordpress/ -type d -exec chmod 755 {} \;

Then change the group ownership of whatever directories you need write access to and give the group write access:

code:

chgrp www-data wp-content/

You'll also need to do wp-content/plugins, wp-content/themes, and any other directories that WP may need write access to. On this same set of directories where you're changing the group, you also need to change the group permissions to allow write access:

code:

chmod g+w wp-content/

What some guys also recommend is simply leaving the user and group alone until you need to upgrade, and then doing this:

code:

chown -R www-data:www-data wordpress/

Do the upgrade, and when the upgrade is done, change the user back to what it was before. Kind of a pain, but it works.

[edit2]
Just wrote about this in much more detail on my site. This should clarify things.

http://www.fangsoft.net/?p=227

Fangs404 fucked around with this message at 09:03 on Mar 7, 2011

# ¿ Mar 7, 2011 04:44

Fangs404: Dec 20, 2004; I time bomb.

JHVH-1 posted:

The point of suphp is that you can isolate users to their sites, and they can only read/write their own files. If you chown it to the user that the web server uses like that then if you have another site and an exploit is able to put any kind of php code on the system it will be able to also modify those files. Its a big help on an environment like cpanel where you are giving accounts to other people and running various websites with code packages you don't have control over. If someone leaves their Joomla out of date or runs crappy scripts that require modifying files a lot it keeps it from spreading elsewhere and getting code injected into all your sites (and thus getting flagged by google).

Another option: add mod_security and set that up with a good ruleset. That will protect the heck out of your site, even if the code stinks and keeps a log.

If you want wordpress performance you run a caching plugin (w3 total cache is my favorite) and combine it with memcache and/or the php apc module.

I understand what suphp does. I think you misunderstood my point. I expressly state that chowning is a very bad solution (read my blog entry). The solution I came up with (selectively chgrping just a few directories) is better than chowning everything (more secure) and better than suphp (much faster).

Most php scripts I use don't actually modify files/directories. They simply edit a database. My WP blog is the exception to that. I don't want/need suphp for every php site I run.

# ¿ Mar 8, 2011 09:28

Fangs404: Dec 20, 2004; I time bomb.

loving Christ, having a VPS is loving amazing. It's awesome to just be able to install whatever the gently caress I need to get poo poo going. I don't know how I ever lived with shared hosting before. And it's loving fast, too (well, Linode is, at least).

gently caress.

# ¿ Mar 9, 2011 10:04

Fangs404: Dec 20, 2004; I time bomb.

Got a question regarding SSL in Apache 2.2. I've got it working just fine, but I'm having an issue that I can't figure out. So I have several sites I'm running all on the same IP. I have VirtualHosts setup for them all, and everything's working just fine. I want one of these VirtualHosts to have SSL enabled, and I want it explicitly disabled for the other sites (basically, an error message should pop up). When I just do a simple SSLEnable on that one VirtualHost, though, all of the sites attempt to use that one VirtualHost's cert and allow SSL. How can I fix this? If this doesn't make sense or you need more info to diagnose, lemme know.

# ¿ Mar 22, 2011 08:52

Fangs404: Dec 20, 2004; I time bomb.

optikalus posted:

You can set your VirtualHost entries like:

<VirtualHost *:80>
</VirtualHost>

and then

<VirtualHost 1.2.3.4:443>
</VirtualHost>

for the SSL host.

The other hosts will still answer on 443 because SSL binds to the IP address (and they'll have an invalid certificate warning), but it should load the default virtualhost (first virtualhost apache finds in the configs).

If you create a virtualhost entry at the top of your list like:

<VirtualHost *:443>
ServerName 1.2.3.4:80
</VirtualHost>

It should load that one since the ServerName doesn't match any valid *:443 vhosts, and you can point it to an error page if you want.

Thanks man. What I wound up doing was just creating a cert for all the sites I run using a wildcard CN.

# ¿ Mar 25, 2011 06:38

Fangs404: Dec 20, 2004; I time bomb.

CampingCarl posted:

Is there a general consensus on where to get domains from?

Everyone will recommend something different. I personally am a big fan of https://www.nearlyfreespeech.net/ due to their simple interface and nice pricing model.

# ¿ Mar 28, 2011 08:34

Fangs404: Dec 20, 2004; I time bomb.

melon cat posted:

Does anyone have any experience with configuring an FTP app like FileZilla for use with a wireless router? I'm have a hell of a time getting FileZilla to work. I've checked every online resource, called Belkin (router manufacturer), and nothing's working. I realize that Wireless Routers can cause problems when it comes to configuring an FTP and it involves Port Forwarding but I'm really lost on what exactly needs to be done.

Is it my router? Are there any particular routers out there that work best with FTPs?

All you have to do is forward port 21. http://portforward.com/ will tell you how - just find your router model and then select FTP on the applications page (if you see a message about PFconfig, hit "skip this advertisement" in the top right corner).

# ¿ May 2, 2011 06:40

Fangs404: Dec 20, 2004; I time bomb.

http://blog.linode.com/2011/06/16/linode-turns-8-disk-space-25/

Linode increased everyone's disk space by 25%! Just resized my main disk image to add another 4gb. Gotta love Linode!

# ¿ Jun 19, 2011 01:11

Fangs404: Dec 20, 2004; I time bomb.

No one's mentioned it, but Linode just recently began rolling out IPv6 support to all its customers. http://blog.linode.com/2011/05/03/linode-launches-native-ipv6-support/ and http://www.linode.com/IPv6/ explains more.

I just spent the last hour or 2 working on getting my sites IPv6-compatible (adding AAAA DNS entries, editing VirtualHosts, etc.), and it seems to have gone smoothly, although I don't really know of a way to test it other than this site which says I pass.

# ¿ Jul 28, 2011 09:56

Fangs404: Dec 20, 2004; I time bomb.

Cross-posting this from http://forum.linode.com/viewtopic.php?t=7699 because you guys might be able to help:

First of all, what I have:
Linode 512
32-bit Ubuntu 10.04 LTS
Apache 2.2.14
PHP 5.3.2
MySQL 5.1.41

I'm running http://americankpopfans.com/ for a couple friends, and they had a huge spike in users a couple days ago, and ever since then, it's been having performance issues. The original issues were due to hitting the swap hard. I realized MaxClients in Apache was way too high (it was at the default of 150), so my settings now look like this:

code:

<IfModule mpm_prefork_module>
    StartServers              1
    MinSpareServers           3
    MaxSpareServers           6
    ServerLimit              15
    MaxClients               15
    MaxRequestsPerChild    3000
</IfModule>

That fixed the swap issue, but now the CPU is going berserk. It looks like this:

It's pretty much all due to MySQL, too. I've been trying to tweak it, but it doesn't seem to be helping. Here are my settings:

code:

key_buffer          = 8M
max_allowed_packet  = 1M
thread_stack        = 128K
thread_cache_size   = 8
max_connections     = 100
table_cache         = 128
query_cache_limit	= 1M
query_cache_size    = 32M

They have a decent amount of traffic, but it's not outrageous by any means (it's on the order of 2000-6000 hits a day). I don't really know what to do right now. I've also installed Quick Cache which is working (view the source on the homepage and scroll to the bottom). For some reason, though, it's still hitting MySQL hard and is loading super slowly. Does anyone have any ideas about what might be causing this?

# ¿ Sep 6, 2011 18:46

Fangs404: Dec 20, 2004; I time bomb.

Bob Morales posted:

What's the output from top look like? Sort the processes by CPU usage (hit 'O' then 'K' after top loads)

Here are a few samples taken several seconds apart:

ClosedBSD posted:

I found this blog here with some options for Wordpress tuning/caching to cut down the number of MySQL queries, I don't know much about wordpress but I doubt your website is updated fast enough that you shouldn't abuse as many caching techniques as possible

http://elliottback.com/wp/why-my-wordpress-site-is-so-much-faster-than-yours/

I already have caching enabled in wp-config.php, and I'm using the Quick Cache plugin (similar to WP-Super Cache). It's weird to me that the database is getting hit at all. Once the popular pages get cached, the database shouldn't be getting hit again for another hour.

Fangs404 fucked around with this message at 20:15 on Sep 6, 2011

# ¿ Sep 6, 2011 20:10

Fangs404: Dec 20, 2004; I time bomb.

Bob Morales posted:

You're still paging like crazy, I bet. Each apache process is using ~60mb and there's 9 of them. You have almost no free RAM in that second screen shot.

Do an 'iostat -m 1' and you'll see how much it's writing every second.

How big is your mysql database? Go along with what ClosedBSD was saying, there are a lot of guides out there for running a VPS in a lower-memory setup. 512MB isn't as bad as 256MB or 128MB but you're running on the edge of hitting swap.

Also change your font in Putty to Consolas or something only scrubs use Courier

code:

Linux 2.6.39.1-linode34 (galactica) 	09/06/11 	_i686_	(4 CPU)

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          27.51    0.00    2.85    1.21    1.25   67.17

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              8.72         0.11         0.02       1699        335
xvdb             37.05         0.10         0.12       1416       1729

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          84.35    0.00    8.18    0.23    6.07    1.17

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              3.00         0.02         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          88.97    0.00    6.71    0.00    4.32    0.00

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          91.83    0.00    7.18    0.00    0.99    0.00

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          90.84    0.00    8.42    0.00    0.74    0.00

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          91.83    0.00    7.18    0.00    0.74    0.25

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          18.27    0.00    1.16    0.00    0.00   80.56

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              2.00         0.00         0.03          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          13.41    0.00    0.42    0.00    0.14   86.03

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          13.02    0.00    1.74    0.00    0.00   85.24

Device:            tps    MB_read/s    MB_wrtn/s    MB_read    MB_wrtn
xvda              0.00         0.00         0.00          0          0
xvdb              0.00         0.00         0.00          0          0

The database isn't huge by any means. ibdata1 is only 18mb.

[edit]
I just changed MaxClients to 10 just to drop memory usage (I now have 135mb free). However, MySQL is still taking up 55-80% CPU, and the site is still loading slowly.

Fangs404 fucked around with this message at 20:29 on Sep 6, 2011

# ¿ Sep 6, 2011 20:22

Fangs404: Dec 20, 2004; I time bomb.

Bob Morales posted:

Okay, doesn't look like you're paging. Can you go into mysql and just do some simple things like 'show databases' and 'select * from articles' and see if happens instantly or not?

I wonder if you have a Wordpress plugin or something that is making the whole site slow. Think back to when the site started acting up, and what the last thing you did was...

Just ran some queries, and they ran almost instantly. I'm starting to wonder if it's a plugin too. Is there a way to log every query MySQL gets or a way to watch them in realtime? If I could see every query it's being issued, I could probably identify the problem.

[edit]
I don't think it's a plugin. They aren't using many, and the ones they are using are really popular and should have any kinks like this worked out:

[edit2]
Using apachetop, I can see that it's getting like 2.5req/sec. I can't see a reason why mysqld should be so tied up.

Fangs404 fucked around with this message at 21:12 on Sep 6, 2011

# ¿ Sep 6, 2011 20:55

Fangs404: Dec 20, 2004; I time bomb.

Bob Morales posted:

It does seem like you're loading a ton of poo poo on the front page.

One thing you could try is disabling all the plugins and re-enable them one by one, it'll either gradually get slower or one will just gently caress the server.

Before that, you might want to look up 'mysql slow query log' and see if there's a certain one or two that are doing all the damage.

Yeah, you're right. It was set to show 25 posts. I changed it to 15.

If I put to site in maintenance mode using this plugin, all of a sudden it's fast. As soon as I go back to live mode, though, it's slow. Disabling all the plugins in live mode doesn't seem to really do anything in regards to helping speed things up.

Now, I'm not so sure it's MySQL. I enabled the slow query log, but the slowest query is taking only like 0.003 seconds. I'm kinda at a loss here. I don't really know what else to check. Is there something in Apache I should be looking at? I think the caching plugin is doing it's job, but it's like Apache is just struggling to get out the data fast enough for each request.

# ¿ Sep 6, 2011 22:02

Fangs404: Dec 20, 2004; I time bomb.

Thanks a ton for the suggestions! I'm writing a paper right now, but once I'm done, I'm gonna start playing around more with this. I'll respond later tonight.

# ¿ Sep 6, 2011 23:36

Fangs404: Dec 20, 2004; I time bomb.

Biowarfare posted:

Your thumbnailing plugin is either terribly designed or I don't know what the gently caress, because while normal images load instantly, anything thumbnailed by it takes upwards of 30 seconds to appear.

e: If you want, i'll take a look at your box and optimize if necesary (free), nothing else to do on a boring night

If you're really willing to, I'd love some help. Shoot me an email - gcfairch@gmail.com.

KeepAlive is off, HostnameLookups is off. Here's the netstat output:

code:

     2 CLOSE_WAIT
     23 ESTABLISHED
      2 FIN_WAIT2
    110 TIME_WAIT

As for MaxClients, everyone on the Linode forums says that for the Linode 512, MaxClients should never exceed around 20 as 512/20 = 25.6mb per client, and even that is probably too much since it leaves no room for MySQL or any other services. I was still hitting the swap with 15 clients occasionally which is why I dropped it down to 10.

Fangs404 fucked around with this message at 02:02 on Sep 7, 2011

# ¿ Sep 7, 2011 01:59

Fangs404: Dec 20, 2004; I time bomb.

ClosedBSD posted:

Might be time to switch to nginx - a little bit more elbow grease involved, but shouldn't be hitting as hard as apache.

http://library.linode.com/lemp-guides
http://library.linode.com/web-applications/cms-guides/wordpress

Yeah, I've been reading about some people suggesting this. I was really hoping I wouldn't have to.

After I restart MySQL or Apache (either one), the site responds really fast for about 30 seconds, and then it resumes responding super slowly. Why would that happen?

# ¿ Sep 7, 2011 04:25

Fangs404: Dec 20, 2004; I time bomb.

I think I just fixed it! W3 Super Cache with page caching, db caching, and object caching all done with APC seems to have solved the problem! It's using virtually no CPU, and I have 150mb of memory free.

Thanks everyone for the help!

code:

ab -t30 -c5 [url]http://americankpopfans.com/[/url]      
This is ApacheBench, Version 2.3 <$Revision: 655654 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, [url]http://www.zeustech.net/[/url]
Licensed to The Apache Software Foundation, [url]http://www.apache.org/[/url]

Benchmarking americankpopfans.com (be patient)
Completed 5000 requests
Finished 9100 requests


Server Software:        Apache/2.2.14
Server Hostname:        americankpopfans.com
Server Port:            80

Document Path:          /
Document Length:        74546 bytes

Concurrency Level:      5
Time taken for tests:   30.005 seconds
Complete requests:      9100
Failed requests:        0
Write errors:           0
Total transferred:      681408215 bytes
HTML transferred:       678368600 bytes
Requests per second:    303.28 [#/sec] (mean)
Time per request:       16.486 [ms] (mean)
Time per request:       3.297 [ms] (mean, across all concurrent requests)
Transfer rate:          22177.66 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.1      0       8
Processing:     7   16  56.0     14    2443
Waiting:        4   15  55.3     12    2405
Total:          7   16  56.0     14    2444

Percentage of the requests served within a certain time (ms)
  50%     14
  66%     16
  75%     18
  80%     19
  90%     23
  95%     26
  98%     30
  99%     33
 100%   2444 (longest request)

gently caress yeah.

Fangs404 fucked around with this message at 05:24 on Sep 7, 2011

# ¿ Sep 7, 2011 04:50

Fangs404: Dec 20, 2004; I time bomb.

Big shoutout to Linode. I had to submit my first support ticket. They responded within 3 minutes of me submitting it, and we had the problem solved in 48 minutes. And it's almost midnight! I didn't know support that good existed.

# ¿ Sep 13, 2011 05:45

Fangs404: Dec 20, 2004; I time bomb.

Linode now has a datacenter in Tokyo, Japan.

# ¿ Sep 23, 2011 18:07

Fangs404: Dec 20, 2004; I time bomb.

Regarding TLDs, what do you guys recommend for personal sites/blogs? I was reading about the .name TLD and think that may be the way to go, but I'm not sure if .com is better (I guess for SEO purposes).

# ¿ Dec 10, 2011 21:01

Fangs404: Dec 20, 2004; I time bomb.

jerkstoresup posted:

I feel I should tip my hat to Linode for their service. "Due to a catastrophic hardware failure, unfortunately all data was lost" for the machine/s my node is on. So they applied a 3-month credit ($60) to my account as compensation. Fortunately for me and my data, I only bought a slice at Linode a few months ago and used it to idle irssi, and I'm still lazily waiting for my Bluehost account to finish before I move my sites to Linode.

They don't discriminate between active/important nodes and seemingly unused ones like mine, which is quite nice. But I will certainly maintain my own backups once it's there, even though I'm sure this is rare for them.

Do you use their automated backup service? I'm wondering if the data would've been recoverable if you had.

I heart Linode as well. Only had to deal with their customer service once, and it was because of a DoS. They helped me diagnose and fix the problem within like 45 minutes of me submitting the ticket.

# ¿ Dec 14, 2011 20:03

Fangs404: Dec 20, 2004; I time bomb.

jerkstoresup posted:

Nope, hadn't setup a backup with them or personally at all at the time (though I will when it matters). They do say that their backup service successfully kept backups for those that had them. Here's the explanation ticket after the initial notice:

Just to reiterate for anyone skimming, I'm sharing this as an endorsement, not a knock against Linode. Even if I had important stuff on the node, I'd be the dumb one for not maintaining backups of any kind, and their response is a good one to me.

Nice, that's good to know. And yeah, I did read it as an endorsement. Comping you 3 months of service is really awesome of them. The reason I asked about the backup service is because I do use it, and I wanted to know if it was effective here. Sounds like it was, so I'll definitely keep using it!

# ¿ Dec 15, 2011 02:39

Fangs404: Dec 20, 2004; I time bomb.

Linode's pretty loving awesome. If your site contains stuff you care about, you ought to consider enabling their backup feature. It's only $5 a month, and it'll give you a little peace of mind.

Fangs404 fucked around with this message at 03:24 on Jan 17, 2012

# ¿ Jan 15, 2012 00:22

Fangs404: Dec 20, 2004; I time bomb.

Ubik posted:

Well, looks like I was dumb to stick with Made2Own when Brutal.net bought them. My sites with them have been down for almost two weeks now, and my last email from them was 9 days ago when they claimed they were under a DDoS attack. Emails since have fallen on deaf ears.

So, I'm pretty much done with these guys. All I have is a website dedicated to my music and brewing, so nothing terribly mission critical other than "don't go down for weeks/months without so much as a refund." Recommendations for something barebones, cheap and reliable are welcome. In the meantime, don't do business with Brutal.net.

Honestly, if you want cheap and reliable, go with Dreamhost. They're always running insane deals like a year's hosting for $10 or something. I was with them for 4 or 5 years before moving, and I hardly ever had any downtime. Support was always super fast to respond, too (I had a response for every ticket I submitted about 30-45 minutes after submitting it).

# ¿ Jan 18, 2012 07:53

Fangs404: Dec 20, 2004; I time bomb.

dvgrhl posted:

Any host that offers yearly hosting for $10 will just potentially put Ubik in the same position. You really do get what you pay for. Sometimes it doesn't bite you, sometimes it does.

I'd agree with this for every host except Dreamhost. They've had their issues in the past, but in general, they're reliable. They've been around long enough to know how this whole shared hosting thing works. Ubik said he doesn't have anything mission-critical. It sounds like he's mainly using it for a file dump or small Wordpress blog, and DH is great for that.

# ¿ Jan 18, 2012 18:30

Fangs404: Dec 20, 2004; I time bomb.

JHVH-1 posted:

I prefer w3 total cache. It does database caching, object caching, javascript/css minify, supports CDNs and varnish.

Yeah, total cache is better. And its use of APC is pretty awesome (and a shitload faster than using the file system).

# ¿ Jan 22, 2012 06:44

Fangs404: Dec 20, 2004; I time bomb.

madkapitolist posted:

Does anyone have any recommendations for a cheap VPS in EUR for downloading torrents? This will be for personal usage and pretty low volume.

Linode has a facility in London, but if all you're using it for is torrenting, maybe a cheaper VPS would be better. What about a seedbox?

# ¿ Feb 9, 2012 03:21

Fangs404: Dec 20, 2004; I time bomb.

thegasman2000 posted:

I am looking for some hosting. Its somewhat different from my regular reselling hosting in that it will be a backend for a large iPhone app were launching. I probably want VPS and some redundancy. Can I get 2 VPS deals and sync them? Could that share the load so I dont rape bandwidth on one account?

Any good deals / Goon Deals?

Linode offers NodeBalancers that do exactly what you want.

# ¿ Apr 25, 2012 06:46

Fangs404: Dec 20, 2004; I time bomb.

fuf posted:

I've been happily hosted with mediatemple for years but I want to start using django and they don't support it on their shared hosting.

Do you think I'll be able to find a host that'll let me use django for less than $20 a month?

It looks like dreamhost offer it but what's the general consensus on dreamhost these days?

e: what about webfaction? Opinions?

Get a VPS like Linode, and you can install whatever you want. Django is part of the reason why I went from MediaTemple to Linode. Once you go VPS and discover how flexible root access is, you'll never go back.

# ¿ May 17, 2012 23:43

Fangs404: Dec 20, 2004; I time bomb.

gmq posted:

How's Mediatemple's shared hosting?

I have a Linode 512 but lately I'm starting to get tired of having to take care of the server, messing with config files to ensure it works under a high load, etc.

I went from MT to a Linode 512, and MT's performance isn't anywhere near what Linode offers. If you want the performance of Linode but without the hassle of an unmanaged VPS, why not get a managed VPS?

# ¿ Jul 15, 2012 16:55

Fangs404: Dec 20, 2004; I time bomb.

Tars Tarkas posted:

I am not sure if disk I/O will be a problem with the hosting, I'd hadn't even heard of it outside of personal computers until I started looking into VPSs. From the sound of it, it might be something that will bottleneck me in the future if I can keep the site growing. The only problem would be if one of my plugins is badly coded and uses up too many resources, but that should be causing performance issues already. That's if I understand hosting disk I/O right as most topics talk about it as I should already know.

What plugins are you using? Are you using one of those plugins that shows related posts? I had some issues with those in the past. A lot of them are horribly coded and just destroy disk I/O/CPU (they rely on complicated SQL queries which is fine if the results are cached, but a lot of the plugins don't cache the results). If you are, try either disabling it or switching to another related posts plugin that properly caches results.

# ¿ Nov 25, 2012 01:49

Fangs404: Dec 20, 2004; I time bomb.

Tars Tarkas posted:

I had the stats part of Jetpack active, and is one of the things I have turned off atm (the only other part of Jetpack I used was Sharedaddy). P3 declared podPress was the biggest problem and that is deactivated as well, though that's how I have my podcast set up and is more important than stats that I can get via statcounter.

WP Super Cache (the wordpress cache my webhost told me to use over others)

Did you actually notice that the stats part of Jetpack was slowing things down? I have it active and haven't noticed it do much at all.

You ought to give W3 Total Cache a try. W3TC has a shitload (overwhelmingly many) of options, but after tweaking it, I got much better performance under W3TC than with WPSC. As a first go, just enable the page cache, database cache, and object cache in the W3TC settings using APC. You might experience better performance.

# ¿ Nov 25, 2012 19:35

Fangs404: Dec 20, 2004; I time bomb.

Linode just increased everyone's storage by 20% for free. This is part of the reason why I continue to stick with them.

# ¿ Dec 15, 2012 01:21

Fangs404: Dec 20, 2004; I time bomb.

McGlockenshire posted:

Time go to see how online filesystem resizing works.

If I don't make it back, tell my wife and kids that I love them.

It's braindead simple and takes like 30 seconds. Shut down server, resize, boot back up.

# ¿ Dec 15, 2012 01:29

Fangs404: Dec 20, 2004; I time bomb.

mmm11105 posted:

What's the hivemind's opinion on mediatemple's (gs) service? I'm shopping for hosting for a client that will be featured on a national tv program shortly after the new site launches, so it needs to be able to hand one hell of a traffic spike.

I had their gs, and while it was definitely stable, it wasn't particularly fast, and it took them a forever to upgrade things. Things might be different now, but they promised something called the cluster server since like '08 that was supposed to replace the grid service. It still hasn't happened. It wasn't until mid-2009 when they finally offered MySQL 5 on gs. Back when I moved off of them (2011), they were still running PostgreSQL 7. I just looked it up, and they don't even offer PostgreSQL on gs anymore.

Anyway, if you're gonna have any sort of significant load and you care at all about modern software, just get a VPS. I switched to Linode, and I've been much happier. It's snappy (a lot of people think it's one of the fastest VPS places out there), and because you have root, you can install whatever software you want.

# ¿ Jan 8, 2013 04:03

Fangs404: Dec 20, 2004; I time bomb.

Milkie Galore posted:

I've never understood how MT managed to get a reputation among media/creative types for being good quality or whatever. They've always struck me as style-over-substance garbage for idiots with more money than sense, and pretty much everything I've read about them confirms this.

Very much so. I will say this - the one time I contacted their customer service, it was good. They responded very fast. But their whole development system or team or whatever needs to be reevaluated. Promising a major upgrade to the gs service and never delivering after 5 years is absolutely loving unacceptable. And jesus christ, waiting until mid-2009 to push MySQL 5? gently caress that noise.

# ¿ Jan 8, 2013 04:23

Adbot: ADBOT LOVES YOU

# ¿ Apr 28, 2024 12:56

Fangs404: Dec 20, 2004; I time bomb.

I was fortunate enough to sign up with the free Google Apps before they shut it down (my domain is ((first_initial)(last_name).com). I'm having a hard time deciding what email prefix(es) to use to best organize my email. What are some commonly used personal email address prefixes? I'd at least like to make a distinction between personal contacts (emails from friends/family) and account/money-related stuff (emails from Amazon, PayPal, Facebook, etc.).

Some I've thought about for the personal contacts:

- first_name@domain.com
- mail@domain.com
- contact@domain.com

For other stuff:

- accounts@domain.com
- money@domain.com
- finance@domain.com
- junk@domain.com

So what are the most commonly used prefixes for different situations in order to best organize/segment emails?

# ¿ Feb 2, 2013 02:51

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Hosting/Webhosting MEGATHREAD

«‹›2 »