Geovario move - Anyone else having responsiveness issues?

[Bryansix]

If your ISP is lame and ignores TTL then it should follow that you should use a different DNS provider like google at 8.8.8.8 and 8.8.4.4 or Open DNS at 208.67.222.222 and 208.67.220.220. Note you would have to manually set this in your home router or in your computer's local area network connection.

[Nose Nuggets]

the site isn't "slow", its hard to find.

when your browser actually locates the page, the page content should load normally. its the actual finding of the page that is taking time.

if you are still having this issue, you can manually change your DNS provider to google (8.8.8.8) and should see and immediate improvement.

[AragornElessar86]

See, my problem is tapatalk only. Mobile browsers all work fine, PC browsers all work fine, Tapatalk only work sometimes, and with no identifiable pattern. Anyone else having this problem and/or know a fix? I should add that I even uninstalled and reinstalled and still no dice.

[Kodemonkey]

This should be pretty much resolved at this point, if your Internet Provider is ignoring TTL they shouldn't have it set much higher than 3 days.

If you are still having problems in WinXP/Vista/7 :

click start > run > and type "cmd"

This will drop you to a dos shell.

Type : "ipconfig /flushdns" (without the quotes)

This will kill your local DNS cache and let it pull in the younger (and much hotter ) DNS from your DNS server.

[Kodemonkey]

See, my problem is tapatalk only. Mobile browsers all work fine, PC browsers all work fine, Tapatalk only work sometimes, and with no identifiable pattern. Anyone else having this problem and/or know a fix? I should add that I even uninstalled and reinstalled and still no dice.

What phone is it running on? if it is an iphone just do a hard reset (press and hold the power and home button until you see the pretty little apple logo). Actually, a hard reset should work for pretty much any phone.

[AragornElessar86]

What phone is it running on? if it is an iphone just do a hard reset (press and hold the power and home button until you see the pretty little apple logo). Actually, a hard reset should work for pretty much any phone.

Android 4. I'll try the hard reset.

[tonyxcom]

DNS doesn't explain the inconsistant page loading and performance.

If this was a DNS problem related to our ISP we wouldn't load the site at all. The fact that we can resolve the site but page loads take 30secs/never one hour and then act normally the next do not indicate a DNS issue.

Telling people to change their DNS servers is not a solution to this problem and doesn't always result in faster name resolution for most users since the 3rd party DNS is off of their ISP's network.

[tonyxcom]

DNS doesn't explain the inconsistant page loading and performance.

If this was a DNS problem related to our ISP we wouldn't load the site at all. The fact that we can resolve the site but page loads take 30secs/never one hour and then act normally the next do not indicate a DNS issue.

Telling people to change their DNS servers is not a solution to this problem and doesn't always result in faster name resolution for most users since the 3rd party DNS is off of their ISP's network.

[tonyxcom]

Wow, a double tap NOT related to a DNS issue.

[Kestryll]

Okay, time for an update, this is late in coming as I've been dealing with some offline stuff. Nothing critical just my Wife heading out to Texas for a week for work. It took up most of my weekend however as she's never done this and hasn't flown since before 9/11.
I'm a panicky, overprotective wreck.

Anyway..
As you know we moved from one data center to another, at the same time the IP address of the forum was changed. Complicating this, based on IP tracking, the new IP address is one that had previously been in use somewhere outside the U.S.
This made for some interesting DNS propagation issues.

We had hoped this was the whole of the issue but apparently not.
Over the last few days I've been getting some database error notices and then today I have been inundated with 'more than 'max_user_connections' active connections' errors.
Several thousand tonight alone.
Each of those emails is someone clicking on a page of the forum and it not loading.

I have contacted GeoVario and they are aware of the issues and are working on resolving them.

My take on it is that when we were moved to a new data center it wasn't our server being physically moved but our data moved to a different server and database server.
One of the unique things about Calguns is the volume that goes on here, this forum moves more like an enormous chat room and less like a forum and we never purge old posts or threads so it's all there to be read through if someone wanted to.
I'm guessing getting something as established and large as this that moves at this speed up and running on a new machine is somewhat tricky.

From what I've seen the DNS issues seem to have resolved themselves and now we are left with getting things running smoothly again on a new server.
It may take a bit of time and tweaking but I know we'll be back to being as reliable and fast as we were before the move, the Techs at GeoVario haven't let us down yet and I'm sure they won't now.

As I get updates and more info I'll post it and let you all know what is going on.
Thanks everyone for being patient, I know it's not easy sometimes.

[Kestryll]

I told you I'd keep you all updated as I got new info, this just came in today.
I'd summarize but that's what cut and Paste is for!

As you may have noticed ( if you are on LAMP series plan ; Windows ASP plans were not affected ) - we've been having general slow downs during peak traffic periods; since late last week and certainly this past weekend.

We are happy to report that we have greatly alleviated the issues causing this by bringing additional servers into the LAMP Web Clusters to solve the problem. With additional servers still coming in a couple days to solve any last outliers.

As you also probably know we like to be a full disclosure company. So I'd like to explain how this all happened?

As with most things in life - it was not a single issue:

In an effort to provide greater levels of scalability and service, we decided several months ago - to move to a much larger data-center with more competitive pricing ( and more space available ).

This meant that to continue to serve you and your customers during our transition; we needed to buy another set of servers. However; with our low end servers costing thousands; and our larger machines costing tens of thousands - we couldn't just go out and buy a complete set of servers. Instead it was decided that buying the minimum # needed to run for a short period would be best - then as soon as everything was moved - we would quickly ship all the servers from the old data-center and add the additional servers into rotation - effectively giving us at least another 50% capacity to our overall server farms when all was done.

So last week we moved the final set of larger customers out of our older data-center; our jobs were done! We started the process of shutdown, unracking and boxing everything in the old data-center.

Then literally as we were driving the truck to pickup servers at the old data-center... 2 critical servers in the new data-center failed. Our systems are designed to be fault tolerant; so the affected websites traffic simply stopped going to those 2 servers, and all traffic went to the remaining servers.

Unfortunately there was simply not enough capacity on the remaining servers to handle the peak traffic load. Which is why starting at 6pm or so - your website became very slow or unresponsive as our remaining servers did their best to keep up - generally by about 10pm the number of requests would reduce and your sites would return to full speed.

As of last night - we finished the first round of racking, reloading software and adding additional servers into rotation for the Web Clusters; and we aren't done yet - even more servers are being prepared today although at this point we are waiting for our provider to finish installing another power circuit.

I do apologize for the issues this data-center migration has caused, but we look forward to the fact that we are in a much better facility - with a much better method to scale and handle your new websites, or your existing websites with more and more customers and clients.

Thank you for your patience with us during this time