Tweet
A bit of a background:
My site is currently on a 1GB VPS hosted nby Liquid Web. I've been with them for about three years, was always happy with their support, and recommended them to others. However, two weeks ago, I started getting massive load spikes; into the high double digit range. No amount of httpd or my.cnf tweaks or database optimizations would stop it. The spikes came with no increase in traffic; in fact, traffic has been steady for the past two years. I beefed up my robots.txt and .htaccess files to stop obnoxious spiders, along with Configserver firewall for WHM (blocking all traffic from Russia, China, Ukraine, Belarus, and the Philippines), and still it didn't help. (I'm having problems installing the script eva2000 now requires to gather stats for optimization requests, and requests using the old cut-and-paste-the-results-from-shell-commands method are no longer answered.
)
So, I'm wondering ... do I need a 2GB server for a vBulletin "big board" (~600,000 posts since 1996) that gets 100-120 visitors in a 15 minute period peak, 70-100 through the workday, 25 to 70 off-peak/weekends? Liquid Web support says "yes". Still, eight years ago, when my message board was at its peak, with about twice the activity as today, it did okay on shared hosting. Is there anything more I should be asking of them?
If I decide to leave Liquid Web, I'd be looking for a VPS from a host with US- or Canada-based servers and tech support, who isn't a reseller, and who doesn't knowingly host spammers, spam proxies, and the like. I'd prefer a host based in the Great Lakes region, Northeastern US, or southern Ontario. I like the Liquid Web model; their servers aren't in a city that's remote from their offices.
My site is currently on a 1GB VPS hosted nby Liquid Web. I've been with them for about three years, was always happy with their support, and recommended them to others. However, two weeks ago, I started getting massive load spikes; into the high double digit range. No amount of httpd or my.cnf tweaks or database optimizations would stop it. The spikes came with no increase in traffic; in fact, traffic has been steady for the past two years. I beefed up my robots.txt and .htaccess files to stop obnoxious spiders, along with Configserver firewall for WHM (blocking all traffic from Russia, China, Ukraine, Belarus, and the Philippines), and still it didn't help. (I'm having problems installing the script eva2000 now requires to gather stats for optimization requests, and requests using the old cut-and-paste-the-results-from-shell-commands method are no longer answered.
)So, I'm wondering ... do I need a 2GB server for a vBulletin "big board" (~600,000 posts since 1996) that gets 100-120 visitors in a 15 minute period peak, 70-100 through the workday, 25 to 70 off-peak/weekends? Liquid Web support says "yes". Still, eight years ago, when my message board was at its peak, with about twice the activity as today, it did okay on shared hosting. Is there anything more I should be asking of them?
If I decide to leave Liquid Web, I'd be looking for a VPS from a host with US- or Canada-based servers and tech support, who isn't a reseller, and who doesn't knowingly host spammers, spam proxies, and the like. I'd prefer a host based in the Great Lakes region, Northeastern US, or southern Ontario. I like the Liquid Web model; their servers aren't in a city that's remote from their offices.
The real solution would be to find the real reason for increase in load. Is there any plugins that you recently installed / any changes you made to the configuration? Is there any pattern to the load spikes, like it happens at same time every day?
Comment