I'm grumpy about this situation, so please bear with me.
I've had a VPS with JaguarPC since they bought out DEHE; I was with DEHE for well over a year before that.
My VPS has been on one of the old DEHE-configured nodes (UL10) since the merger/buyout. For the most part things were fine, until the last few months, where performance had gotten a bit sluggish. Several support tickets transpired. On multiple occasions I inquired to make sure that *I* wasn't the one causing resource problems, and I was told that while my VPS was one of the higher resource-using VPS's in the node, there were still several others higher than mine. Never at any point did I receive any notification that my VPS was using too many resources.
Fast forward to March 2008. On March 20, my VPS was moved to one of the new boxes. Xeon E5405 Quad-Core, etc., etc. Things are wonderful! Everything is MUCH faster. I go about my business...
Yesterday (5 days after the move) I receive the following message:
Needless to say, this caught me completely off guard.Your vps "vps.xxxxxx.com" on hardware node yyyy.nocdirect.com is utilizing over 20% of the CPU of the node constantly which is causing load problems for the entire hardware node. After carefully watching the server hardware consumption at the time when the load increases and server performance decreases, we have identified that your
vps is the one causing problems.
We run an equal CPU share VPS environment, which most other providers do as well but your VPS is constantly consuming 20% of the CPU of this Intel(R) Xeon(R) E5405 Quad core processor and is not acceptable resource usage for a vps.
The high cpu usage and load on your vps is mainly from the high traffic, apache usage of your account 'zzzzzz'.
This email is a courtesy notice to provide you an option to upgrade your account to your own Dedicated Server. It is advised that you upgrade to a dedicated server as soon as possible, otherwise we will be left with no choice but to disable the server temporarily to stabilize the hardware node if the situation does not change.
I wasn't hogging resources on the older slower node, but immediately upon movement to a newer fast box with more resources, I'm now red-flagged? There's been no spike in traffic.
So I inquire as to this conundrum:
To which I receive the following:Something is wrong then. I was just moved to this brand new node a few days ago. I was on UL10 for well over a year, and my VPS was not the highest using VPS on that node. Now I'm on a much faster machine and all of a sudden my VPS is using too many resources even though I have seen no increase in traffic since the move?
Never once was it suggested that I was using too many resources on the old node but now on a faster machine with more resources I am using too much?
Something is misconfigured. Please recheck what was done during the move of my VPS.
I respond:I just checked the statistics for ul10 and you were the highest resource user on that node as well and you were using 40-50% of the cpu on that node until you were moved to this new node. You need to move to a dedicated server for the kind of traffic/resource usage of your account zzzzzz, which I believe is your main domain domainname.com.
At that point I got a bit irate and requested that the issue be forwarded to management to figure out what the heck is going on. Am still awaiting a response.Now hold on a sec. If you look back through the support tickets, I specifically asked if I was the highest user on the node if my use was causing problems, and I was repeatedly told NO. No one has EVER suggested that my usage was causing a problem, but now all of a sudden, less than a week of being moved to a new node I'm being told I need to upgrade to a service where you offer no specials whatsoever, that is going to cost 3x to 4x what I'm paying now???
What irks me the most is that nothing is EVER brought to my attention about my VPS using too many resources before the node move. Only after the fact, after the threat of shutting down my VPS unless I upgrade to a dedicated server, is anything mentioned about resource usage on the old node. Now I'm magically using 40-50% of node resources.
Me being the cynical, suspicious type, immediately suspects extortion because I'm still grandfathered in under DEHE's old pricing model. That was all fine and dandy as long as I was taking up space on an old DEHE node that no one else wanted to use, but now that I'm on one of the "good servers", Abracadabra! Presto! Resource Hog!
The timing is far too suspicious.
The domain in question gets about 5,000 unique visitors a day, pushes about 100GB of bandwidth per month. I've seen sites mentioned here with far higher statistics.
At any rate, the idea of going to a hosting package that is 3x to 4x the cost of what I am currently paying is... well... insane.
Question: The implication in many of the posts here is that a semidedicated account actually has more resources at its disposal than a VPS, the downside being that it's a single domain, not multiple ones. Since it appears that (according to support), one domain is the problem, would there be any advantage to my moving the high-traffic domain to a semi-dedicated account, while retaining the VPS for all my other domains?
How many semidedicated accounts are there per server? Buying multiple semidedicated accounts (or the equivalent resources) would still be cheaper than the dedicated server options offered by Jaguar.
P.S. I'm willing to work with Jag on this; I just don't like the way it's been handled and I don't trust the information I'm receiving at the moment... something's rotten in Denmark. Am I missing the big picture here? Thanks.


LinkBack URL
About LinkBacks



Reply With Quote

Bookmarks