IT Guy here. It looks like there was a small window of activity between the time we ran our web server backup and the time when the forums were put into a read-only mode. The database server was copied after the site was frozen, so we're good on all of the posting activity.
Back to the MIA images. I believe that there are only 12 images that were missed in the migration. I will move those over in the morning. That should make us whole again.
One more thing. For those who might be wondering what the update was about, we moved this and a number of our other web sites from our own data center to the Amazon cloud. From your perspective, you should see a site that is far more responsive in the States. We will soon enable something called Content Delivery Network that will cache web content much closer to your individual locations. Our friends outside the US will see much better page load times.
I have just started working for a company that uses Amazon to host its websites. It used to be housed in Singapore, but now it's in the newer Sydney centre. Quite different to when I last worked in IT.
Update - I restored the missing files a few minutes ago. The .spin and .zip files are appearing correctly and the inline images are viewable if you click on them. It looks like we need to generate a few thumbnail images to correct that. I'm looking into that now.
@Cluso99 - We have been in AWS since 2014. Honestly, Once we saw how fast they were innovating, we stopped buying server hardware.It's almost unreal to be able to deploy an entire network/server environment from a JSON file.
From your perspective, you should see a site that is far more responsive in the States. We will soon enable something called Content Delivery Network that will cache web content much closer to your individual locations. Our friends outside the US will see much better page load times.
Eagerly waiting for this step to be achieved/implemented ASAP.
As for now, everything is behaving like a lazy worm, from my perspective.
From your perspective, you should see a site that is far more responsive in the States. We will soon enable something called Content Delivery Network that will cache web content much closer to your individual locations. Our friends outside the US will see much better page load times.
Eagerly waiting for this step to be achieved/implemented ASAP.
As for now, everything is behaving like a lazy worm, from my perspective.
Henrique
+1
The response time of the forums are now super SSSSSLLLLOOOOWWW.
The "Bad Gateway" happens when the web server goes missing. It happened this morning when I restarted the web service. For the Linux gear-heads, the CPU load was over 30 when I arrived this morning. It should be around .08.
Whichever you've did with that new server, it's running like a tachyon now! On par with Google servers, from my humble perspective!!!!
The solely concern that remains here, is about the early refered "Content Delivery Network that will cache web content";
If that "cache" contains all the up and downstream contents we where grabbing from the forums server(s), what will happen with PMs, for example? They'll be also cached? Without https://??
The cache will manage outbound content only, all inbound content is still delivered to the web server and not cached at the endpoints. I sent you a PM concerning the https issue.
The cache will manage outbound content only, all inbound content is still delivered to the web server and not cached at the endpoints. I sent you a PM concerning the https issue.
Hi Jim Ewald
Excellent work! Thanks by your explanations. Now, it's crystal clear to me (https://)!
We're still working out the bugs on the new network configuration. Let me run through the list again to be sure all the sites are routing correctly. There is a default setting that should take you to www.parallax.com if the site you wanted to reach has gone offline.
Turning off HTTP without a redirect would send the visitor the the www.parallax.com site. Ideally, the request sent to http://site.parallax.com would generate a 302 response to the https://site.parallax.com. The client browser would then re-request at the new URL. But first, I need to slap SSL certs on the remaining sites. Obex is up next.
Comments
-Phil
-Phil
IT is now working to resolve.
Back to the MIA images. I believe that there are only 12 images that were missed in the migration. I will move those over in the morning. That should make us whole again.
One more thing. For those who might be wondering what the update was about, we moved this and a number of our other web sites from our own data center to the Amazon cloud. From your perspective, you should see a site that is far more responsive in the States. We will soon enable something called Content Delivery Network that will cache web content much closer to your individual locations. Our friends outside the US will see much better page load times.
@Cluso99 - We have been in AWS since 2014. Honestly, Once we saw how fast they were innovating, we stopped buying server hardware.It's almost unreal to be able to deploy an entire network/server environment from a JSON file.
Getting back to fixing the thumbnail images...
Have you ever received the PM I've sent to you?
About forum's security certificates?
Henrique
It's kind of pointless anyway as the forum is also served up using plain HTTP.
Thus login data can be viewn.
Henrique
Eagerly waiting for this step to be achieved/implemented ASAP.
As for now, everything is behaving like a lazy worm, from my perspective.
Henrique
The response time of the forums are now super SSSSSLLLLOOOOWWW.
15 seconds or so to even getting the first response after hitting a link. Then some more seconds to load everything.
Which is an order of magnitude or more slower than my web servers running on google cloud instances on the east coast.
I don't think CDN will address that slowdown.
Got "502 Bad gateway" for a while. Guess the server crashed or some such.
Now it's much much faster
+1 in the US.
Whichever you've did with that new server, it's running like a tachyon now! On par with Google servers, from my humble perspective!!!!
The solely concern that remains here, is about the early refered "Content Delivery Network that will cache web content";
If that "cache" contains all the up and downstream contents we where grabbing from the forums server(s), what will happen with PMs, for example? They'll be also cached? Without https://??
Henrique
Hi Jim Ewald
Excellent work! Thanks by your explanations. Now, it's crystal clear to me (https://)!
PM answered!
Henrique
Using Chrome or Firefox:
Visiting https://www.parallax.com/ gets us to the right place with a valid certificate.
OK.
Selecting Support->Forums gets us to http://www.parallax.com/
ERROR: Not using HTTPS link.
Entering the secure URL: https://forums.parallax.com/ gets us to the learn.parallax.com (Even if it says https://forums.parallax.com/ in the URL bar). The certificate is flagged as invalid with a red "Not secure" message.
ERROR - Landed on the wrong page. Even the wrong site!
ERROR - Certificate is not valid.
This is somewhat of a mess. And not secure!
"There are issues with the site's certificate chain (net::ERR_CERT_COMMON_NAME_INVALID)."
When I hit that link.
Edit: OK. Panic over. Restarting my Chrome persuaded it to pick up the valid cert.
What about disabling the non-secure connection possibility?