Down servers - response time

[Bunchadogs]

Curious what the expected reply time is to a trouble ticket for a down server?

I started out on Live Chat and was told to create a ticket.

Now nearly 4 hours later, no reply, problems still exist, affecting multiple domains on Web18 / Cluster1

No updates on Facebook either to indicate a system issue...

That's a lot of down time with no communication from Jodo!!!

 

[Bunchadogs]

After five hours I get this response:

Can you please check now and update us if still continue to see a similar problem.
​

The server error log is literally full of 500 or 503 errors - shouldn't tech support be able to monitor that without asking ME the question?

UpTrends.com and DownForEveryoneOrJustMe.com all show outages or 30+second http timeout errors....

Very disappointing day - top it off with a local hard drive / NAS crashing AND the cloud backup showing "corrupt" backups everyday for 90 days and this is a really bad week.

 

[Bunchadogs]

Just an example:
75.137.40.210 - - [24/May/2017:13:24:44 -0500] "GET /_scan.php HTTP/1.1" 200
141.255.30.67 - - [24/May/2017:13:24:35 -0500] "GET /wp-login.php HTTP/1.1" 404
75.137.40.210 - - [24/May/2017:13:24:56 -0500] "GET / HTTP/1.1" 503
141.255.30.67 - - [24/May/2017:13:25:11 -0500] "GET /wp-login.php HTTP/1.1" 503
172.76.154.174 - - [24/May/2017:13:16:46 -0500] "GET /newsroom/ HTTP/1.1" 500 -
74.40.27.206 - - [24/May/2017:13:16:06 -0500] "GET / HTTP/1.1" 500 -
66.249.73.235 - - [24/May/2017:13:16:07 -0500] "GET / HTTP/1.1" 500 -
75.137.40.210 - - [24/May/2017:13:24:32 -0500] "GET /music HTTP/1.1" 200


When it does respond it's not pretty - I've watched several where the Time To First Byte is measured in MINUTES - 18.25 seconds is "fast"
Request sent
0.26 ms
Waiting (TTFB)
18.25 s
Content Download
3.49 ms

 

[zardiw]

Curious what the expected reply time is to a trouble ticket for a down server?

I started out on Live Chat and was told to create a ticket.

Now nearly 4 hours later, no reply, problems still exist, affecting multiple domains on Web18 / Cluster1

No updates on Facebook either to indicate a system issue...

That's a lot of down time with no communication from Jodo!!!

Join the Club...........there IS no Live Chat........and support has been 'working' on a MySql Server issue for 5 HOURS now......lol..........

I think this is the answer ........after over 10 YEARS with Jodohost:

http://www.bing.com/search?q=reliab...-20&sk=&cvid=2768130AEB9A4078B05E1725A569E4B0

z

 

[djfenom]

I feel your pain, we've had terrible performance on web16 and web18 for about 3 weeks now.

All I keep getting is we're working on it and it will be resolved in the next few days, but nothing seems to be getting done!

I opened up a ticket 3 days ago and made it clear I'm looking for a proper estimate to tell my clients and think some sort of compensation is in order, but I've had nothing back.

Something is going on, it's like they're avoiding us!

 

[Stephen]

You aren't being 'avoided', no far from it, we're working very hard to fix any issues. We're making some huge network and server changes to upgrade the entire infrastructure and part of that is with some major DDOS protection upgrades. As part of these upgrades there are times that traffic may be offloaded from our network and this introduces a latency between Web/SQL servers that impact some peoples code and sites more than others. We're working to make them accessible in alternative means directly for internal usage so that they do not every his the offload systems.

However, many people have used our proxy service meant for OUTSIDE management via sql query tools for managing, on their internal strings and this is making the issues even worse for them, that was never intended or provided for web server side use, only for allowing clients to get in as we dont allow direct communications to the SQL servers. I'm not saying this is your exact case,but I've seen it such for dozens of strings. If your connection string is using one ending in '201,PORT' or specifying a non default port in the string otherwise, it is a major part of the problem.

 

[Bunchadogs]

So how does that relate to simple CMS sites like WordPress, Drupal, and Joomla?

Are they suddenly configured incorrectly?
That's an honest question - I've always plugged in the name (mysql##.xyz.com) for the relevant MySQL server (instead of localhost), created my DB user, and let the install script finish up everything else.

If JodoHost is not ideal for people simply looking to host a "normal" website maybe you should tell the users?
Or - if JodoHost is not meant for custom database hosted services, maybe you should notify those users?

Instead we're all down...I guess?

 

[Stephen]

Instead we're all down...I guess?

ummm...No. It is up and going fine with a few exceptions we've had, but we are already doing checks/verifies for those situations now to prevent from happening again.

Your complaint as speed. I said and told the reason why the speed could be reduced, and how we are addressing it. It works just fine using the name, but for the BEST SPEED with the network changes to help with DDOS matters, it is going to need some changes to be optimal.

 

[Bunchadogs]

So you're saying SQL servers are fine...

Check your own server logs, they are FULL of 500 and 503 errors!
"It is up and going fine..."
" It works just fine using the name.."

I guess that's only the SQL servers, what about web servers?

What is the ETA on those fixes?

Right now load times are measured in seconds or MINUTES and that is IF you don't get a 500 / 503 errors.

THAT is NOT "fine" - and that is what we REALLY need answers for!

I sent support numerous examples of downtime monitors showing errors or http timeouts after 30 seconds - from around the world.
This is not a "local" connectivity issue - the server has issues, it's not "fine"

Here's a simple Joomla site - 2.5 MINUTES Time to First Byte! This is not an "extreme" example, I can probably get this any time....

Look at these and that's only 30 minute interval testing!