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.
BuyShared + NodePing - should I be concerned?
raindog308
Administrator, Veteran
in General
Every morning, I wake up and find NodePing alerts about downtime on BuyShared.
I had @NodePing (which does an http check) set at 1 minute and backed it down to 3 minutes, which cleaned up a lot. But I still see some every day.
I am not checking my web site at 5am so I don't know what the real impact is, but NodePing is just doing a GET so I assume it didn't answer.
I've never run NodePing against shared hosting so I don't know if this is typical...thoughts? I'm aware of the price I'm paying and this isn't really a complaint, just asking what others' experience is what and what's reasonable.
Here's a typical day:
And some history:
Comments
I think those are just service restarts. That's pretty much what I see with my monitoring on BuyShared as well.
It's possible some of the node ping IPs might have gotten blocked by their firewall
I've similliar situation
most of time got alert from monitoring, sometimes from my friend. downtime only few minute but often
It was the same with my LU Buyshared account. Downtime almost every day, UptimeRobot eMail every few hours. The main reason why I left them.
The interesting thing about HTTP check is that failures can be caused on multiple levels. I always suggest coupling them with ping checks to start with. If you get HTTP failiures alone and not ping failures, then I would start doing two things:
Also, a monitor will often fail an HTTP check if it has a low timeout and the web application is performing poorly. That can be the result of an unoptimized application, hitting up against CloudLinux limits, or a poorly performing web/db server at the host. Narrowing that down may be difficult, Fran can probably help you rule out all but the application itself.
Probably they are getting blocked in the firewall aswell, I somehow doubt that BuyVM/BuyShared would have all these outages.
Almost same as yours with nixstat and uptimerobot but so far I didnt find any problem compare to random check directly on my some website when alert coming.
Only one I found till now so far confirm on @Francisco when they have problem with Fiberhub power outage.
I think its mostly like service restart. I still believe they have great horse to race till now since I get in touch fast respon when I need Frantech team so In my opinion all the HW resource they have, their always keep in open eye and care for it.
I experienced many outages (restarts?) as well, most of them only lasted a few minutes.
I'm sure BuyVM/Shared will pop along, however I remember them stating in the past it was due to the large amount of domains / vhost records during a Apache restart when someone adds a new account / domain.
Didnt buyshared run some fancy script that allowed every user to get his own ip, and hence required a deamon restart whenever a user/domain was added?
Might be those short downtimes are whats being caught in your monitor.
I have reason to believe it is Apache gracefully restarting - it simply means someone has just created an account
I have accounts with buy shares in Luxembourg and Las Vegas. Monitoring detects last down time 20 days ago.
Have you opened a support ticket with them?
I think that's how Francisco explained the situation a while ago. Every time a user adds a domain, you'll see a minute or two of downtime.
Sounds like they're doing it wrong if that's the case. You can reload Apache configurations without a server restart.
We use litespeed which has had more than a few patches for systemd my own monitoring showed it restarting every night when upcp runs, but it's not often I get anything during the day.
Open a ticket and I can take a look. I know I had to downgrade to an older version of litespeed a while back to get SSI working again.
Francisco
Along with a ping to the same IP, I would suggest a DNS check on the FQDN as well. Sometimes the lookup is the 'slow' part of the HTTP request.
We've had conversations about timeouts with quite a few shared hosting providers and it usually turns out to be server load issues due to cron tasks like log rotation or database re-indexing.
If you'd like the check to ignore flapping like this, I recommend decreasing the 'sensitivity' and increasing the 'threshold' on the check. That will give the service more time to recover and avoid 'down' notifications being sent. If the notifications are not actionable, they're noise and noise will increase the chance you'll ignore an important 'down' notification.
Ive had the same but from a Wordpress monitor - Jetpack with a somewhat similar thing but it doesn't monitor per minute, more like per 5mins or so I believe?
"Hi ,
Good news — your site http://-------- is back up!
Your total downtime was 27 mins, but your site was up again as of Saturday, April 16, 10:53 am.
If it goes down again at some point (we hope it doesn’t!), we’ll be in touch.
Cheers,
The Jetpack Team"
Is it just me or does this thread make you want to purchase a BuyShared account?
I don't think this only happens to BuyShared, I have a Stablehost shared hosting account and using uptimeRobot to monitor and I get emails once a month at least.
However they come back quickly, normally within 5 minutes.
So I guess it is just a common issue on shared servers.
No, it's not. 100% Uptime with my Prometeus Shared Account and 100% Uptime with my dirtcheap Hostmantis account.
Two exemplary days from my short time with BuyShared LU:
I'm not sure why anyone else would want hosting that has crazy amounts of downtime notifications but go for it! It's only a few dollars.
I get the same NodePing alerts on my BuyShared service also, but I also get them on my own cPanel server but much less frequently and only during high loads so I'm assuming the same thing is happening on the BuyShared servers (if you go browse your website right when the alert comes in you'll see the sites are online still, it's only the NodePing checks timing out). I'll tweak my monitoring tonight to see if I can reduce the number of false alarms I get a day.
Here's my NodePing results for BuyShared since I started monitoring it:
http://pastebin.com/pPuxbQ9D
And my NodePing results for SD's cPanel server for the same months:
http://pastebin.com/c3bf2yrQ
You can see BuyShared is showing a lot of downtime but I am positive that the majority of them are false alarms.
At least in my case, the website was indeed unreachable when the alert came in. Not a question of timeouts.
Yeah! Call a mod because @raindog308 is clearly shilling for BuyShared!
:-)
That was my experience too (and in Vegas, not LU).
I've always been really surprised at how positive people are about BuyShared on LET because it was far and away the least reliable hosting I've ever used. Daily downtime isn't normal!
If you got charts/etc send them over. I've got no alerts for any other nodes and 03's the only one that's rebooting litespeed at night during upcp's. As I said, we're having to use an older version because SSI is broken in the latest litespeed builds.
LU's ones were because of networkmanager and the DC botching a router upgrade. Once the cut the cables over Network Manager dropped all of the user IP's. Since I disabled that and such LU's been perfectly fine.
Francisco
yea, flagged! @raindog308 @Amitz @jarland