The company that I work for is looking to buy a SAN to use as a storage backbone and for VM's and as part of our distributed disk infrastructure. Since cost is an issue, we are looking at IScsi running on a gig network. We have looked at DellMD3000i, and IBM DS300s anybody have any experience with either. Any pros/cons. Also any other hardware recommendations in a similar cost.
View 0 Repliesi'm using RHEL 4, apache 1.3.37, php 4.4.6 and xcache 1.20. Now i'm looking into memcached and running it on more than one computer. So i got few questions:
1) Is it smart to run web/db on one server with 64mb memcached and remote box which is around 150ms from web/db with 2gb of allocated memory for memcached?
2) I saw if I want to use memcached i need to load in php memcache also. Since I already got xcache, will those two produce any products if they will work simultaneously.
I mostly need memcached to save mysql queries, and i'm happy how xcache caches dynamic pages.
I hope somebody here is using memcached. QPS are around 95.
I intend to store and serve files that is around 1-200mb from servers that is physically separate. I wish to store 2 or 3 copies of each file as a redundancy solution. When the file is requested I want to serve a direct downloadable link to the file. I am looking for logic to distribute load, and to detect if a server is down. I also whish to redistribute the files if a server is taken out of the pool..
Is there any open/free systems for this I can use?
I'm looking into setting up a server-monitoring network: small globally distributed VPS servers all monitoring another main server (currently, just one server - but I would like it to be scalable).
When it notices it can't connect to a particular service, it should check to see if the other monitoring stations also can't connect - then a "master server" should send out an email + sms alert to the designated person for that server if all servers cannot connect.
Does such an open source piece of software exist? Or do I have to start writing my own...?
if you have a dedicated server, is it possible to let anyone become part of your dedi server using a software d/l?
eg: You put an app on your site for users to download,users download and run it, so now their internet connections become part of ur dedicated server, so when people access ur website, some data is downloaded from ur dedi server, and some data and cpu usage is used from the user's net connections who downloaded and are running the app.
Any commercial/open source software like this that lets website visitors become part of ur server?
We've been seeing sluggish performance on our mail gateways, and so I started doing some digging in the logs. It looks like we are filling up with messages like:
2007-05-16 12:22:16 Connection from [xx.xx.xx.xx] refused: too many connections
We have our max connections set to 20 (total, not host-specific) in exim4. So I started tailing the logs, and sure enough, we are getting bombarded with requests to randomstring@ourdomain.com coming from all over the map. The requests are getting denied of course, but that doesn't help the connection issue since they are consuming all of them, preventing real mail (for the most part) from getting through.
What is the proper way to deal with something like this? I could certainly just up the max connections value from 20 to 40 or 50 or whatever, but I'm not sure what kind of performance impact that will have on the rest of the traffic going through our gateways.
Since the spam attempts are coming from all over the place, it doesn't seem like I can just firewall out a few addresses and be done with it.
This particular rack is a cluster of web and database servers behind two gateway boxes, which handle the mail traffic (so this problem is on the gateways, the actual mail server itself sits behind the gateways and never actually sees these fake emails).
My server has a small SAS disk(about 73G), if I use 90% diskspace of it, is it good idea, will it harm the physical HDD?
View 6 Replies View RelatedI am looking for better disk performance. Due to the tight budget, I have to choose one of following options as my disk choice:
2 SATAII disk w/RAID0, 7200rpm, 32M cache for each disk
1 SAS disk, 15000rpm, 16M cache.
which one will be better and how better if other things(hardware & OS) are same?
can a Xen disk image be converted to a diskpartition?
Someone is asking whether I can host his disk image at his current host, which he is leaving for poor I/O (wonder why that would be ). I can host a diskimage, but I don't like diskimages (slow, and 100GB isn't very 'comfortable' either). Is there any way out there to convert a disk image into a normal partition?
I use apache with CentOS VPS hosting for my blog. I only host one blog in this VPS account. I have 1.5GB RAM and I have 7, 500 page preview per day. My page loading time is 2-3 seconds (according to the pingdom tool).
I want to know what is the best performance (faster web page loading) W3 Total cache option for VPS hosting blog. Currently I use Disk to enhance for page cache and database cache for disk.
Under Service Status, the "Disk (dev) 100 % ", what should I do?
FreeBSD with cPanel installed.
My server load is high, i checked and see everuthing is ok.
I think my sata disk cannot support my hard disk traffic.
Is it posible to check wich file used more hard disk traffic? (rpm speed)
how can i reduce this value
it was at 70% but after i did wget for abig backup and its failed i get this value in dev/sda5
Quote:
[root@xx]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/sda5 4.0G 3.6G 196M 95% /
/dev/sdb1 135G 111G 18G 87% /backup
/dev/sda1 198M 40M 148M 22% /boot
none 3.0G 0 3.0G 0% /dev/shm
/dev/sda8 76G 37G 36G 51% /home
/dev/sda7 2.0G 37M 1.9G 2% /tmp
/dev/sda3 25G 6.0G 18G 26% /usr
/dev/sda2 25G 20G 4.4G 82% /var
/tmp 2.0G 37M 1.9G 2% /var/tmp
alos
Quote:
[root@xxxx]# du -h --max-depth=1
1.8M ./tmp
8.0K ./opt
1.2M ./namedOLD
728K ./named
84K ./profiles
18G ./lib
423M ./cpanel
4.0K ./portsentry
28K ./empty
16K ./lost+found
288K ./run
672M ./cache
8.0K ./local
8.0K ./preserve
28K ./db
217M ./log
1.2M ./www
40K ./lock
16K ./crash
8.0K ./nis
8.0K ./net-snmp
24K ./yp
12K ./account
100K ./netenberg
170M ./spool
19G .
As per this [url]tutorial to set up disk quotas u need the /Home in the fstab file but how ever all i can find on my server is this:
LABEL=/ / ext3 defaults 1 1
LABEL=/disc2 /disc2 ext3 defaults 1 2
tmpfs /dev/shm tmpfs defaults 0 0
devpts /dev/pts devpts gid=5,mode=620 0 0
sysfs /sys sysfs defaults 0 0
proc /proc proc defaults 0 0
LABEL=SWAP-sdb2 swap swap defaults 0 0
LABEL=SWAP-sda2 swap swap defaults 0 0
How do i go about setting quota for my users
shared hosting env?
/etc/sysctl.conf:
vm.dirty_background_ratio = 20
vm.dirty_ratio = 60
blockdev --setra 16384 /dev/sda
I have a VPS that has hard limits for Disk Inodes at 500,000. I am currently over 430,000.
I am only using 24GB of 40GB of disk space.
How do I find what is gobbling up disk inodes? How can I find unused, abusive or unnecessary disk inodes and remove them?
# df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/vzfs 500000 430553 69447 87% /
/dev/simfs 500000 430553 69447 87% /tmp
/dev/simfs 500000 430553 69447 87% /var/tmp
none 2049640 95 2049545 1% /dev
From the Disk I/O performance is it better
1) to have main PHP file with 10 includes
2) all 11 files as one file
3) the difference is not big
Suppose
a) a low traffic site
b) a high traffice site
I have a server with cpanel/WHM. The file system says I am using 26GB of my 40GB:
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 40G 26G 14G 66% /
none 257M 0 257M 0% /dev/shm
/usr/tmpDSK 243M 4.1M 226M 2% /tmp
/tmp 243M 4.1M 226M 2% /var/tmp
however the total of my accounts is less than 5GB.
How can I find out where all the other disk space is being used?
partitionning a server disk.
I have a 76 gb disk which is partitioned this way :
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sdb5 10317828 319000 9474712 4% /
/dev/sdb1 101086 24171 71696 26% /boot
none 3114756 0 3114756 0% /dev/shm
/dev/sdb7 36384624 17595080 16941268 51% /home
/dev/sdb8 1035660 34732 948320 4% /tmp
/dev/sdb3 10317860 4959788 4833952 51% /usr
/dev/sdb2 10317860 4847048 4946692 50% /var
/tmp 1035660 34732 948320 4% /var/tmp
I would like to create the same kind of partition on another server, but the disk is 750 gb and I need /var to be at least 30 gb big.
How big must the other partitions be ?
Regarding "none" ---> /dev/shm,
I don't understand very well. Do I have to create this partition myself ?
How do I decrease my disk inodes usage?
View 6 Replies View RelatedGood day All
the Disk /dev/sda5 (/var) 96 % Full
What I do
i Have 2 vps From Swvps and HostForWeb
When Finish Transfer Old VPS( HostForWeb ) To New VPS (Swvps.com) and See Disk Usage
Swvps=> 13 gig Usage
HostForWeb=> 23 gig space!
I've got a server in a local colo facility. both the facility and my server are slowly falling apart. Rather than investing in hardware and then shipping it to some location and hoping it doesn't break, I've decided to investigate VPS.
What I am finding is puzzling me. Why do providers charge so much for disk space? My current old p3-733 has a 30gb drive, I could deal with about 15gb but with some 'flex' room (i.e. a virtual drive for uncompressing files or some such). It's a personal server for me and a couple of friends, it doesn't get much traffic, I don't need a heck of a lot of bandwidth, nor anything flashy, just Ubuntu 8.04. I see places offering plenty of traffic for a decent price, yet little in the way of storage. I just don't get it.
A year ago I was here wanting to trade my class C for a dedicated server. I realize now that it would be pretty difficult for providers to do this, because if I wanted to yank my class c they'd have to renumber. But if anyone is still interested, let me know.
I've decided I want cPanel installed on my iWeb dedicated box. They informed me it will require a reinstall. Before they do this I wanted to image the disk and download the image to my PC. Is this possible to do via SSH? I'd prefer an open-source free solution.
View 11 Replies View RelatedI have a site with a lot of data (almost 20-30gigs). I want to pkgacct the site to transfer to another server but the harddisk gets full and the pkgacct stops working.
Which files and folders can I delete on the server without touching critical CPanel files? I want to delete as many as folders so that I can have enough space for the packaging process.
i having an issue with FreeBSD...
I have a Raid5 setup and working fine, I just put 2 new 18gig drives for raid1.
Now i see in dmesg this.
amrd0: <LSILogic MegaRAID logical drive> on amr0
amrd0: 104193MB (213387264 sectors) RAID 5 (optimal)
amrd1: <LSILogic MegaRAID logical drive> on amr0
amrd1: 17365MB (35563520 sectors) RAID 1 (optimal)
In bold is the 'new' array.
in fdisk i get....
fdisk /dev/amrd1
******* Working on device /dev/amrd1 *******
parameters extracted from in-core disklabel are:
cylinders=2213 heads=255 sectors/track=63 (16065 blks/cyl)
Figures below won't work with BIOS for partitions not in cyl 1
parameters to be used for BIOS calculations are:
cylinders=2213 heads=255 sectors/track=63 (16065 blks/cyl)
fdisk: invalid fdisk partition table found
Media sector size is 512
Warning: BIOS sector numbering starts with sector 1
Information from DOS bootblock is:
The data for partition 1 is:
sysid 165 (0xa5),(FreeBSD/NetBSD/386BSD)
start 63, size 35551782 (17359 Meg), flag 80 (active)
beg: cyl 0/ head 1/ sector 1;
end: cyl 164/ head 254/ sector 63
The data for partition 2 is:
<UNUSED>
The data for partition 3 is:
<UNUSED>
The data for partition 4 is:
<UNUSED>
I get write errors when i use sysinstall. As for the raid status. Its fine.
Logical Drive : 1( Adapter: 0 ): Status: OPTIMAL
---------------------------------------------------
SpanDepth :01 RaidLevel: 1 RdAhead : Adaptive Cache: CachedIo
StripSz :064KB Stripes : 2 WrPolicy: WriteThru
Logical Drive 1 : SpanLevel_0 Disks
Chnl Target StartBlock Blocks Physical Target Status
---- ------ ---------- ------ ----------------------
0 00 0x00000000 0x021ea800 ONLINE
0 03 0x00000000 0x021ea800 ONLINE
One of the things people seem to bring up a lot is disk IO Performance.
Why? Because theres little you can do about a customer being stupid and creating a disk swap nightmare.
There is however something you can do to reduce the impact across your clients, Have a separate raid array for swap space.
This does 2 things, it splits some of the Disk IO across 2 arrays, but more important it reduces the affect someone overusing there swap will have on the ones that are not.
Just my quick 10 cents for the day.
One of our server is very slow and cpu and memory usag is fine. Also, there isn't any problem on network, I thought the problem with disk and I got the following iostat result. Is it normal?
# iostat
Linux 2.4.21-37.0.1.EL (domain.com) 08/20/2007
avg-cpu: %user %nice %sys %iowait %idle
6.66 1.87 1.50 8.22 81.75
Disk sda3 (/) 100 %
I can't access via Shell to manage them, can do anything with WHM cause I receive error about no diskspace left.
How else I can remove files?
If console, is there any default url or port to access? Since they didnt gave me this info and suppor t is "somewhere...