Anyone else seeing major slowness when trying to log into the cluster 1 control panel? It takes anywhere from 15-60 seconds for us to log in, from multiple internet connections, multiple computers, multiple browsers (we’ve tested a bunch of ways to connect to rule out it being an issue outside of the CP itself).. I have a ticket in, probably my 5th or 6th one, and JH does “something” that speeds it up for about 1 day, then it goes back to slowness. So I’m curious if others are having that same issues..
We will soon announce Control Panel maintenance in this weekend to resolve such slowness problem.
Was this ever fixed?
Today the control panel is REALLY slow - it used to be just the Reseller Control Panel - user accounts were fine.
Today I tried a number of user accounts, and they are all incredibly slow.
This is on Cluster 1 - it this an issue with old accounts?
Is it possible to migrate?
The slowness is almost unusable at this point.
This was fixed, but today again it is responding slow.
We are looking into it and will fix it soon.
I’ve noticed it slow every day this week.
Sometimes it errors out with a 500 and 503 errors.
Also today horrible performance. Is the CL#1 control panel EVER performing well ???
We have announced control panel maintenance tomorrow to resolve this slowness issue.
The slowness is really due to how Hsphere is designed. To show quotas, FTP info, disk usage etc it has to query every server in the cluster to open panels. We’re going to be announcing some major changes soon, SiteStudio will be killed in an OS upgrade (because its over 10 years old and not updated to support modern OS), this will help some as it will reduce some services requirements. We’re also moving to Intel SSD based storage networks for the control panel services with the OS upgrades, other servers are being upgraded as well. These changes will help a lot when completed.
Ok thanks!
It seems speed in general is getting to be a problem on Cluster1
Numerous database driven sites experience random errors during the day/peak usage - these are sites with minimal traffic of their own.
Word Press and Joomla sites are the worst, but even generic PHP and MySQL sites experience slow or failed queries…
The appeal of having a central control panel for a re-seller is almost completely lost when it takes 5-20 seconds to navigate.
- We’ve announced End of Service on SiteStudio and will begin focusing on changes to CP servers soon.
- Speed issues are being addressed, and will likely require editing the DB server IP address if you are having issues with these. I’ve explained how we’re making some major changes in network to help address ddos problems, and this requires some offloading of network to cloud at times to scrub bad traffic, among other changes that can dynamically occur. This makes a latency in the cluster servers that we are making workarounds for and once we have a complete list we’ll post it to customer only.
"This makes a latency in the cluster servers… "
Unfortunately when users experience Time To First Byte hitting 10 seconds - and sometimes up to 2 minutes! - that is an OUTAGE, not downtime.
That may not match your terminology, but that’s the realty.
As a re-seller we’re the front line - you may experience our frustrated trouble tickets, but WE’RE getting that times ten or a hundred!
The added latency is on a matter of 10ms-15ms each way, it doesn’t impact everyone. In fact we used to have MSSQL and Web servers in NJ and Miami at the same time, and really didn’t have a huge amount of complaints, and then again in Texas and Miami, and there were a few then, but nothing like has been since we started this. We’re working as quickly as we can to make things faster than ever, with a DDOS protected network that won’t see the same problems we’ve had over the last 2 years and growing worse as DDOS attacks amp up and get more widespread and easy to launch from IoT and high speed malware infected home connections.
We’ve spend many tens of thousands in this last quarter to upgrade everything, and we’re nearing completion and full rollout. Once it is completed everything should stabilize and be a lot faster, SSD arrays, SAS drives, SAN backends, uniform hardware deployments and more.
We still need an answer for the “big picture” - sites are basically down, sites that have been stable for months or years with no code changes.
Our only recourse is to look to JodoHost for explanations and solutions - and so far it’s been pretty vague.
there’s no one standard big picture answer, we ARE working on upgrading everything end to end from network to servers it will be done by mid June. We’re working our tails off to get it done and make things whole, but cloud offloading is going to be a possibility at any point in time due to the nature of DDOS attacks today, so moving to a local connection string just for your website to communicate to the database server is a big help, in the long run, and there are multiple providers that have required special connections for web servers over the years and some that do today if they arent running on ‘localhost’ for all services.
Are you aware the problems have been consistent all week? These are not “short” periods of “slow” or “latency”…
I guess our concern is “cloud offloading” seems like it’s not a “possibility” it’s how things are running 24/7!
If this is true?
I just tested a fresh install of WordPress, no data, no plugins, just the default 4.7.5 package.
New MySQL database, setup in the Control panel, connecting via URL and user/password - nothing weird, just a “standard” install
This is on Win18 - Cluster 1:
Nobody will wait 25.5 seconds to see if a website will load!
That’s “down” - and that is what we have been dealing with all week.
One more test, 12.9 s TTFB
Third test, 54.88s TTFB
Fourth test, 22.5s TTFB
That covers a 20 minute period of time now - that 12 second time is pretty quick…
(edit)Here’s another one one hour later:
17.16 TTFB - not “bad”, but then 37.74 SECONDS to load a 79.2KB jpg and 14.74 SECONDS to load a 281.2KB jpg