New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
And I'm 2 for 2:
Me:
Voxility:
>
BRB, I'm off to go buy a lotto ticket.
Sorry about that. I know it was stressful. I'll see where I can improve communication more since I know some people were trying to get on the site.
Francisco
Can confirm looks all fixed;
Everything is ok now?
For 5 - 6 hours or so. I documented the replies I got from them above.
I'll work out an RFO with Aldryic a little later, going to be a tad bit hard to expand on what they said though.
Francisco
Wow this was almost like another Delimiter. J/k
Not all all. Delimiter would have given out some coupon codes and pop-tarts by now.
Hah. Remember, this was just filtering that was bugging out in LV. The rest of LV, all of the BuyShared nodes, etc, were perfectly happy.
So... I need to send @jarland an industrial sized barrel of lube? Well if you say so....
Frnacisco
Any provider that struggles with their own name is to be avoided..
How the fuck did I manage that?
Francisco
Maybe need some quick nap?
Haha
I'm actually off to bed in a few. Just going over my hardware list one more time for LV.
Francisco
did I miss something, or all of my nixstat and uptimerobot have wrong analysis. Anyway I still like how @Francisco bring the horse to win the race...
I guess you mean the pony? BuyVM has ponies... and PonyPower always wins
@TheOnlyDK This is the reason I do not allow Unlimited domains. Downtime ;XD
https://www.lowendtalk.com/discussion/comment/1624625/#Comment_1624625 ;P
If your upstream provider is having issues, I don't think disallowing unlimited domains can help in any way.
We do not face this type of issue, our server is not over-flooded ;P
Companies without customers generally do not have load issues on their servers.
By mistake domain was expired, It was a fault of domain registrar. Ok Its acceptable but now upstream provider is culprit ;P
Next time they will claim that server was down due to President of America
1. Domain Expired - Domain Registrar was responsible.
2. 5 Hour downtime - Upstream provider is responsible.
3. I am waiting for next excuse
Would be interested to check it out.
And you'll see it too. I've seen @Francisco joke about "startup mentality" a few times here but the truth is, whether or not he realizes it, he shares some of that mentality. It's easy to set up WHMCS + SolusVM. If you set it up and leave it alone, you'll be praised for your stability for a long time. It's easy to not make mistakes when you're not willing to take risks.
On the other hand, @Francisco has found himself having to explain a lot of outages over the years that have come as a result of Murphy, as well as being the result of creative ideas and experiments in bettering his service.
So while you may praise yourself for the ability to leave your servers idle and notice that they continue to run without intervention (or many customers, for that matter), Fran focuses on growth and building out the product. The result of his tinkering through the years can be see in Stallion today.
I'll take downtime with Fran any day over uptime with some host that boasts uptime purely due to laziness, that will eventually fall when they actually do have to roll the dice some day.
The person who tries the least fails the least number of times. They also succeed the least number of times.
I've warned you of the dangers of sobriety in the past, and now hopefully your eyes are opened.
This will be my quote of this day. Thanks to bring it up here. "Its easy to be a man but not so easy to being a Man"
Do you have a filtered IP? The issue was only with the filtering ranges. If you don't use the DDOS filtering in LV then you wouldn't have seen anything at all
The TL;DR is that Voxility merged a bad ACL that blocked GRE/IPIP/any sort of tunneling traffic on at the very least the switch we're uplinked on. It took them a while to get it fixed since their own rule distribution system kept reinserting the bad rule.
I'm not even sure what to say for Dewlance. I'm sure there's an autoboot joke somewhere but i'm more confused than anything else.
Thanks @jarland, the lubes on the way.
Francisco
Production server is not for experiment
If anyone do experiment on production server then I want to suggest their customers to keep away from them. If you want to do your experiment then do it on test server.
Lot of dumb and beginners do experiment on their production server. If you want to learn then learn on test server and for improving service you can do it on test server and then apply it on your production server.
Example: If you want to replace Apache with Nginx to improve your service then first check that Nginix can handle 50 or 100 customers or not. If you will do this like dumb on your production server then of-course server will crash and you will face 5 hour+ downtime. You will be need to revert to Apache and change lot of setting.
This is the reason I do not recommended new company because of they think that production server is for experiment.
@DewlanceVPS - I think you're mixing something up here. This wasn't a case of tinkering and something went mental. This was Voxility committing an ACL incorrectly and it mixing up the sequence order.
Francisco
Leave @Francisco alone please.
Better take off them glass slippers before you break something @DewlanceVPS
Hah! I don't mind, I'm just a bit confused and think he misread where in the chain the issue was. The OP included BuyShared for some reason even though it would've only been the public facing page, we don't include filtering on any of the plans so no one was affected by it.
We offered SLA to any customers that aren't behind on their bills and have given short break downs of what happened and our communication with Voxility. It was literally sentence at a time though so there's not a lot of meat on the bone to flesh out to something more useful.
Francisco
Lot of dumb and beginners do experiment on their production server. If you want to learn then learn on test server and for improving service you can do it on test server and then apply it on your production server.
You're an idiot. You are misunderstood adding features to his service overall to a single server, like the ones you rent...
Also your example was daft. People know Apache is about as stable as Greek's economy at taking load. Nginx would be better.
If it takes 5 hours to get a web server online after a configuration issue, check your auto-boot, or IQ...
I believe they use LiteSpeed on shared hosting, and while it was just an example, it goes to show you don't know what you're on about.
They thoroughly experiment features before putting them online. No leaks poor performance from their Anycast when it was brought live, and I engineered Anycast routes for 1+ year. Not easy with an uneven AS-PATH.
This is why I would avoid a wanna-be host who trolls around with poor ESL skills, resulting in a fail post as seen above.
While I think Aldyric has customer service skills on par with an Iranian prison, that's personal. BuyVM is solid, and has features others like yourself could only dream of.
@Kris - tips hat
Lets let this die off. I While I appreciate the kind words the issue is long since resolved and it isn't doing anything but slapping some random people around.
Thank you all
Francisco