cottonwolf
Regular Member
- Jan 20, 2015
- 469
- 242
I'll post this question here, because the programming section is largely ignored.
I've got 2 digitalocean servers on 512mb ram each. Each server had serverpilot.io agent installed on it, so it manages both servers. The only customisation I've done was fail2ban install and mod, and some ssh hardening. No caching is done on the servers that I know of.
Serverpilot sets up the webservers with apache 2.4.x and nginx latest. Nginx serves as reverse proxy for apache.
That's what's happening to me on 2 digitalocean servers is:
Server A -> piwik on domainone.tld/piwik. No SSL, so that should be done as well.
Server B -> wordpress on domaintwo.tld. NO SSL, it's an empty amazon aff. wannabe site.
wordpress gets slowned by a lot because it's loading shit from server a. it's probably because of lack of caching.
I'll post speed test pictures here in a minute:
Running speedtests on pingdom, it seems that either their dns is shit or digitalocean name servers take a fucking long time to load.
That's most of google's speed test feedback: http://pastebin.com/szCAj53W
Any knowledgeable dev/marketer willing to share some feedback?
Where should I start caching?
This is pingdom's result from Server A's piwik javascript:
Response Headers 200
Date cut GMTContent-EncodinggzipLast-Modified Date Cut GMTServernginxETagWTF
VaryAccept-EncodingContent-Typeapplication/javascriptTransfer-EncodingchunkedConnectionkeep-alive
Request Headers
Pragmano-cacheAccept-Encodinggzip, deflate, sdchHostdomainone.tld
Accept-Languageen-US,en;q=0.8User-AgentMozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36Accept*/*Refererhttp://domaintwo.tld/
Connectionkeep-aliveCache-Controlno-cache
That's the webpagetest.org speedtest from yesterday http://pastebin.com/g5ke6yFM .
It seems that external dns lookups take a whole lot of damn time.
Please help a peasant.
I've got 2 digitalocean servers on 512mb ram each. Each server had serverpilot.io agent installed on it, so it manages both servers. The only customisation I've done was fail2ban install and mod, and some ssh hardening. No caching is done on the servers that I know of.
Serverpilot sets up the webservers with apache 2.4.x and nginx latest. Nginx serves as reverse proxy for apache.
That's what's happening to me on 2 digitalocean servers is:
Server A -> piwik on domainone.tld/piwik. No SSL, so that should be done as well.
Server B -> wordpress on domaintwo.tld. NO SSL, it's an empty amazon aff. wannabe site.
wordpress gets slowned by a lot because it's loading shit from server a. it's probably because of lack of caching.
I'll post speed test pictures here in a minute:
Running speedtests on pingdom, it seems that either their dns is shit or digitalocean name servers take a fucking long time to load.
That's most of google's speed test feedback: http://pastebin.com/szCAj53W
Any knowledgeable dev/marketer willing to share some feedback?
Where should I start caching?
This is pingdom's result from Server A's piwik javascript:
Response Headers 200
Date cut GMTContent-EncodinggzipLast-Modified Date Cut GMTServernginxETagWTF
VaryAccept-EncodingContent-Typeapplication/javascriptTransfer-EncodingchunkedConnectionkeep-alive
Request Headers
Pragmano-cacheAccept-Encodinggzip, deflate, sdchHostdomainone.tld
Accept-Languageen-US,en;q=0.8User-AgentMozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36Accept*/*Refererhttp://domaintwo.tld/
Connectionkeep-aliveCache-Controlno-cache
That's the webpagetest.org speedtest from yesterday http://pastebin.com/g5ke6yFM .
It seems that external dns lookups take a whole lot of damn time.
Please help a peasant.