Are CRON backups (scheduled from within DA) included in the bandwidth total?

Q

quadium

Verified User
Joined
Sep 29, 2007
Messages
44
Are CRON backups (scheduled from within DA) included in the server bandwidth total?

We started sending some backups off-site... and somehow our in-DA data transfer is 150GB for Nov and 50GB for December thus far, which puts us around 200GB for the month.

But...

The RRDTool graph from our co-lo provider says we are at 850G. Not sure if it's Bits or Bytes, but the company obviously assumed bytes when billing us... $3.00/GBx 650.89GB = $1,952.67. Anybody have any idea which it is? is G bytes, or bits?

What the? DA says 150GB for Nov and 51GB for Dec... either they are out to lunch or those backups aren't being properly accounted for. We normally do between 15-65GB of traffic on any given month... so the fact that we hit 150GB should suggest that the backups are being included.

Anybody have any idea? If I'm going to contest this ridiculous charge and I am...(not to mention $3.00/GB is INSANE). I need to cover my bases so that I can prove to them either they are mis-reading their chart totals or they have a serious problem with their configuration... unless DA isn't properly accounting for data-transfer (bandwidth usage), that is.

Thanks in advance...
 

Attachments

  • data_transfer.jpg
    data_transfer.jpg
    248.3 KB · Views: 153
In addition, we checked our incoming bandwidth from our provider who was receiving the in-bound backups.... around 120GB for the same period.

Hmm... Not looking good for the co-lo... any other explanation here?
 
I suggest you run MRTG on your own server (locally against your eth port) to see your real usage.

If you find it's the same as your datacenter's, then you've got a problem on your server; if you find it's lower, then you've got an argument with your provider. Also double check with your backup provider; if they're sure of the data, and you're using more than they're getting, then the data is going elsewhere. Have you been hacked? Possibly, if you're really using the data.

I don't know if DirectAdmin counts that bandwidth or not; perhaps DirectAdmin staff will reply to the thread and tell us.

Note that your graph shows you using about 14 mbps at 95th percentile usage. If you bought that bandidth from us you'd be paying about $500 to $700 per month.

Edit: And if you bought it from a local T-1 provider you might pay as much as $6,000 per month.

Jeff
 
So my guess is that they are reading the G as bytes... not bits. And in that case would explain the huge differential.

In theory if we transferred the 30GB of data off our server, every single night for 30 days... that sort of data transfer is theoretically possible.

30days x30GB = 900GB

We are backing up to our office, and our provider here only registered 120GB in the same period. (Our normal usage, plus incoming backups.)

Not to mention DA registered approx 150GB, which when compared to normal, non-backup bandwidth, then cross-referenced with the incoming provider makes sense... the anomaly being our colo provider.... Which I assume is them misreading the graphs.

If that's the case, they may well have been over-billing a lot of people. They did just switch to this graph about a year ago, and had a massive issue on deployment that caused everybody's graphs to be highly inaccurate.

Sigh. I used to have PRTG running on that port but the box that had it crashed and I haven't got around to replacing it just yet.
 
About being hacked, we have hardened our server against Brute Force Attacks and I'm notified if someone manages to get shell access.... which I'd get the provider to pull the plug before they do any real damage. I also use fairly tough passwords with symbols and numbers and capitals and all that fancy, pain-to-type stuff. :)
 
You must log in or register to reply here.
Back
Top