email notifactions

[2webmaster]

Hi,

Following accounts experienced issues in the past 1 hours

================================================== ========================
| Username | Domain | Name | CPU | IO | Memory | EP | NPROC |
================================================== ========================
| username |domain name | --- | 0 | 0 | 27.5 MB | 0 | 0 |
================================================== ========================

Are the items shown at time of issue? Because their is no limit being it according to info above. And if the data sent is not at time of issue. what steps do I need to take to make that appear correct?

Thanks
Frank

[sqh]

Completely useless notifications. Really strange that this has not been fixed yet.

- Scott

[ryan]

Since theres zero documentation on email notifications and I feel that they are absolutely critical as clients hitting their resource limits often blame the host for being slow without knowing its their website causing the problems Im wondering what settings everyone is using in regards to frequency and faults triggered.

I don want to send emails too frequently as that would annoy the customer so I was thinking one per day would work but how many faults do you suggest be the minimum to send an email to the client?

[henri]

These emails are less than useless since they dont actually provide any useful information to the end-user or the admin.

At best, the end-user will login to their cpanel and click the Resource Usage link. At worst, theyll log a support ticket and start badgering the helpdesk to explain what is the problem, how it isnt possible their tiny little wordpress website isnt exceeding anything, etc. and then we must try to prove them wrong with no helpful information.

[kevin]

Can a developer elaborate on this? There has to be some obvious reason why this report is failing to produce anything useful. If not how in the world can it be turned off?

[sqh]

I find it incredibly frustrating that such a needed service (notification of faults) is not working. If you want to stay in the hosting business, you need to provide good customer service. We cannot do this, without timely and accurate and useful fault notifications.

- Scott

[kevin]

I agree with Scott and I find it rather rude that a dev has not made a comment here or offered any assistance to resolve this obvious problem. Not at all a professional way to conduct business!

[cloudlinux]

You can find the settings within the Options tab of the cPanel / WHM login
I have mine set to :
send an eMail to admin only
once per hour
at least one fault

I receive mine at xx:13 every hour
If there were no faults in the previous hour I do not receive an eMail

Whilst I agree this notification is a little misleading, login to the clients account (fr om the provided Domain Name) and you can determine wh ere the issue resides
However, this can take time as you may need to look at the raw access logs and the latest visitor list

It can be quicker to simply leave the LVE Manager in view and watch for specific accounts that remain visible and do not disappear

Now for the misleading part
For each account you set the Memory Limit
This is the number of 4KB pages, not the actual amount of memory
Depending on the packages, I have set 256, 512, 1024, and 2048
This equates to 1MB, 2MB, 4MB, and 8MB respectively
I then receive notifications which exceed these limits
You just need to remember to convert the number of pages into MB
These limits may appear incredibly low when the server has a limit of 2,048MB but this pretty much spreads the availability across accounts
In turn customers whose websites are regularly reaching the lim it can then be encouraged to upgrade to a more appropriate account, which in turn gives them more access to higher resources

[kevin]

Misleading is not the word I would use to describe this. I get this alert for all accounts telling me there is a resource issue and yet the report does not show any account that is even close to the limits set. The fact is it is sending me this report for no reason what so ever because none of the accounts are misbehaving. Plus when the report is sending

I have set my parameters like yours, will see if that makes a difference. Thanks for the help...

[cloudlinux]

Mine works perfectly well
Therefore this suggests an installation issue
As you can see below, I receive two eMails; one fr om LVE (cloudlinux) and one from LFD (firewall)
This confirms an issue exists with this account and the cause

Recent example of mine, this account has a lim it of 512 (1MB) :

Subject : Hosting account resources exceeded
================================================== ========================
| Username | Domain | Name | CPU | IO | Memory | EP | NPROC |
================================================== ========================
| [username] | [Domain Name] | --- | 0 | 0 | 3.0 MB | 0 | 0 |
================================================== ========================

Subject : lfd on [server name]: Excessive processes running under user [username]
Time: Thu Oct 30 23:03:01 2014 +0000
Account: [username]
Process Count: 13 (Not killed)
User:[username] PID:677783 PPID:677532 Run Time:4(secs) Memory:226912(kb) exe:/usr/bin/php cmd:/usr/bin/php /home/[username]/public_html/forum/posting.php

[sqh]

Marcus, this has nothing to do with LFD. We are only talking about alerts from CloudLinux (which can be sent to the root owner, reseller and/or end customer).

The example you showed confirms what we are saying. It says that "Hosting account resources exceeded" but it does not specify which resource and it does not say by how much.

Your example email says 0 under NPROC. 0 processes? How is that excessive processes?

- Scott

[bogdan.sh]

I believe this happens when Minimum number of faults to notify admin is set to 0, simply it notify you every time how notification period is configured even when there were no faults. Yes that is a bit frustrating, maybe we need some explanation in WHM > CL LVE Manager > Options for this.

[sqh]

No, we have ours set to minimum fault of 1 for admin notifications.

The notifications are accurate... when we receive a notification, I can see that there was a fault for that customer. The problem is the notifications are not very useful. They do not tell you which fault the customer hit. The notifications give you CURRENT data about that customer (usually this is after the fault has occurred and the traffic has settled down), not the HISTORICAL data that triggered the fault.

Someone fr om CloudLinux has explained this in another thread somewh ere. I believe they know about this problem, but just cant seem to find the time to FIX it.

- Scott

[henri]

Hi everyone. I agree 100% that the notification email sent by CloudLinux is less than useless. It does not help the customer at all unless they are already technical people. At best, it just creates questions for them and then they will contact support for help...

Since the built-in notification system in Cloudlinux is not good, I have created my own email notification system. Here is an example of what our customers receive if they exceed any of their LVE limits:

> Your "username" web hosting account exceeded one or more of its resource limits within the last 24 hours:
>
> - Exceeded the CPU limit of 13% of total server CPU. Your website was forced to load slower to reduce its CPU usage.
>
> - Exceeded the maximum of 75 concurrent website connections. Your website was not available 67 times because of this problem.
>
> - Exceeded the maximum of 116 total account processes. Your website was not available 385 times because of this problem.
>
> To view full details about your web hosting accounts resource usage, including the time of each incident listed above, please click the link below and log into your cpanel hosting control panel, then click the "Resource Usage" link under the "Logs and Statistics" section.
>
> http://domainname.com/cpanel
>
> If your account is regularly exceeding its available resources, please consider upgrading to a higher level hosting plan that includes more resources. If you have any questions or need help with anything, just reply to this email and let us know.

I think this notification is MUCH better than what CloudLinux has, because it includes very specific details about what limit was exceeded, what is the limit, and by how much it was exceeded.

I have already provided the script I am using to create this notification to CloudLinux developers, and I am disappointed they have not bothered to "fix" the useless notification system that is built-in. It shows how much the company is operated by very smart technical people, but have not too much business skill or understanding of what end-user hosting customers actually need.

[cloudlinux]

> The example you showed confirms what we are saying. It says that "Hosting account resources exceeded" but it does not specify which resource and it does not say by how much.
>
> Your example email says 0 under NPROC. 0 processes? How is that excessive processes?

I have set several process limits, this includes the Virtual Memory Limit and Physical Memory Limit
As above, you can find these settings in the LVE tab

The settings I have in place are confirmed from the LFD eMail, which then allows me to determine what exactly the cause is, as it lists the files causing the account to be limited