Capture Generators Temporarily Offline

3 posts / 0 new
Last post
g0dxilla's picture
g0dxilla
The Lizard King
Offline
Joined: 04/15/2008
Visit g0dxilla's Website

Our could service provider suspended our account due to excessive swap usage on our capture generators. We are making configuration changes to correct this issue and will have it resolved soon. Expect new request and refresh capture delays. Updates to follow.

puravida's picture
puravida
Jedi Warrior
Offline
Joined: 09/01/2007
Visit puravida's Website

I don't think that a configuration change will fix this issue, and the cloud service provider has been unable to provide enough information to troubleshoot. However, I have launched a couple of varying instances to test configuration changes..

Rackspace's timing was atrocious. They terminated all services at 9:00pm ET, which is after every one of our vendors' personnel have gone home for the day. So we are left to struggle on our own. I managed to get us hobbling, for now, with about 50 generators and may be able to get that up to 75 generators in the next hour. However, we will probably get shut down in the morning. I was once enthralled with Rackspace but now I am appalled at this latest outrage.

After years of paying premium rates (to the tune of 6-figures!) for extreme dedicated servers, it surprises me to be shut down in a matter of minutes, with no real effort to communicate a problem or work to a resolution. We are working diligently to look at options and migrate to a more established cloud provider.

puravida's picture
puravida
Jedi Warrior
Offline
Joined: 09/01/2007
Visit puravida's Website

Rackspace's techs worked with us and brought our account back online, for the time being.

As a result of this situation, I pushed forward with testing on Amazon's EC2 platform. However, their micro instance is the only size that makes sense for our needs and it does not provide enough CPU resource for our generators to achieve maximum efficiency.

So we began to migrate and performance test on Rackspace's next largest instance (46% more expensive for us), which has proven to be much more efficient, reliable, and cost-effective. This size instance overcomes the previous technical issue, so we should be OK now.

This means that we are back to tweaking our auto-scaling algorithm once again, but I have already seen that average capture times are reduced from 20s down to 10s. As we improve the algorithm, we will also see wait times reduced significantly. It appears that the overall, average time to capture will be cut in half, as a result of this upgrade!

As of now, we are officially back on track.

ShrinkTheWeb® (About STW) is another innovation by Neosys Consulting
Contact Us | PagePix Benefits | Learn More | STW Forums | Our Partners | Privacy Policy | Terms of Use

Announcing Javvy, the best crypto exchange and wallet solution (coming soon!)

©2018 ShrinkTheWeb. All rights reserved. ShrinkTheWeb is a registered trademark of ShrinkTheWeb.