Nice Spec !

Right well we have
Intel P4 775 3.4 2Mb
2x 1Gb Crucial 6400 (800Mhz)
2x 300Gb SATA2 Raid 1 Intel On Board

Never used hc on a dual xeon or any dual machine which could be effecting it.

Right whats your disk quota logs like, have you checked to see how much your logs are filling up, 1gb is crazy i dont think anyways had it to that extreme.

Thinking about it i know someone with a dual xeon and hc and they didnt say disk quota is causing them an issue but that doesn¬t mean there using it, will do some more snooping see if i can find out what exact spec they have and if its running.

Not sure if this may effect the disk quota service but how much disk space is being used by your users? get the real amount off your root folder.

I know in the past hc disk quota used to crash every morning for some reason eventually sohail managed to fix it, bit strange but i think the disk quota service is the most issue ridden part of hc as its really hitting home reading real disk info etc. We actually have an issue with rbx at the min whn the service runs 24x7 we get a major major lag but tis not related to memory build up so for now its set to manual until we need to do a backup.

Well, in HC requirements, it¬s never mentioned that dual Xeon server can cause the trouble.

After reducing DiskQuota to work on every 2 hours, I got some space - now I have to restart HCDQ service "only" 2 times a week. I am restarting the service when it consumes 1GB of memory. But what will happen when number of sites increases?

The size of my HCDiskQuota log file is about 200MB per month, and I rotate this file on monthly basis.

As I told, right now, we have about 750 sites, with approximately 300000 files which occupies 8GB hard-disk space and this should not be the problem. All those info I have proceeded to HC Support in January, but I doubt they properly analyzed the problem.

Regarding RBX, we have analyzed how and what it works, and after that we have decided to make and implement our own backup solution, in which we have more confidence, because we know exactly what it works and how is written.

Right i dont think the dual xeon is effecting hc in a negative way.

We dont have the same amount of sites as you, currently using 5gb of disk space, we run the bandwidth scan every 3hrs, then of course disk quotas run at 2am.

Right another question do your sites use a lot of bandwidth, im thinking theres more data to process then it increases the memory issue, but if i were you set a script to restart the service at 5am every day, for now that is.

It depends what do you mean by "lot of bandwidth" - we use 60-70GB per month and I don¬t think it is too much. We have other servers (which does not use HC) which spends 10 times more bandwidth monthly and has no problems.

I understand, that this DiskQuota is very consuming service, but it has nothing to do with memory leak. In proper developed code things like this, simply must not happen.

Anyway, I did not noticed this problem when number of sites on our server was not so big. I think that lots of other users are not aware of the problem because they have less number of sites, and this memory leaking is not so obvious in that circumstances.

I also have another server for testing purposes - that server is not so loaded, but same thing happening there, only on less scale - after each pass of DiskQuota service, quantity of Used Memory is bigger.

Right been watching disk quota since our last reboot each night its grown by about 5-8mb in size, nothing on the scale of you guys and i cant see much difference, you have more sites and data but that pounts to quantity causing the issues

I too am having problems with this same issue.. Running Windows 2003.. I have to restart the quota service every few days.. or it¬ll end up eat GIG¬s of ram.

Its very hard for us to have perfect environment to replicate this issue on our local machines therefore I request you, please contact support dept. at support@hostingcontroller.com with your server access. We will try to find some work around. Hopefully HC7 public beta is releasing next week and I believe you will nt find such problems in it.

I have reported problem more than a year ago. I have opened ticket in January (Ticket:88479). I told HC support then, and several more times in Forum and on ICQ, and I am telling again: our company policy allows access to servers only for employees - so, I am sorry, but I can¬t give you that kind of access. I have sent all other information and files you asked. Your support promised they will solve the problem in "next" Hotfix. That was in January.

On how many different hardware platforms and Windows versions have you tried? I understand that you can¬t replicate each problem, but I can¬t understand that you cannot replicate the problem that many other users have reported, as you can see in this thread (maybe some of them may give you access to their server). Also, I am pretty sure, that a lot of other users are not aware of problem because they host less number of sites.

When we bought a license for HC - it was implied that you will solve problems and fix issues related to your software which obviously you did not do. So, we have no intention to pay for HC7 upgrade, since you did not solve reported problem related to current version of HC for more than a year. How can anybody be sure that same behavior won¬t repeat with HC7?

HC as an update i¬ve been looking at this issue on our box

On first service start its at 2mb within 4-5 days your at 50-70, small compared to everyone else but something still aint quite right.

It¬s a file handles leaking problem. At each bandwidth (each 1 hour for us) and disk (each night) scan it does not close opened file handles.
This is the situation for us after about a week it is running:
HCDiskQuota Service

Right we have new servers bang big memory leak

Over 8 days HCDiskQuota eats 400Mb of Ram thats a large chunk, we have 1gb in each box so we were using 1.4gb all the time

Did you find a fix or work around from hc, i may have to script a manual service restart at 4am or something like that. At least on disk quota restart the lag goes, defo leaking lol

__________________
Chris Daley
Dwebs Ltd Director :: Company No. 05603664 :: Phone No. (UK) 0870 803 4423
www.Dhosting.co.uk - Web Hosting, Domain Registration, Windows 2003 NLB Cluster with HC 6.1!!!
www.Dwebs.ltd.uk - Web Design & Other Services
My views are my own and not those of my company.

My server has about 1500 websites.
I have to restart DiskQuota and Delivery services everyday. I wrote a batch file for this and run it by Scheduled Tasks.
There are two poblems.
1- When these services are run they use huge amount of VMs and do not free the memory.
2- They use a lot of CPUs