Optimum Configuration

I am going to dedicate a server to FMS. I plan to have as
many as 10,000 connections. Does FMS in linux take advantage of
dual processors? Memory? Disk Space? I won't have sophisticated
applications nor am I doing media streaming. This is purely for
socket/shared object use. I'd liek to know if 4GB of memory helps
over 2GB. Is a large amoutn of disk space required and does dual
processors help.
Also, from what I understand, I only need ports 1935 and 80
open correct?
Thanks!

Well I'd say if you have 10,000 connections then no thats not
enough hardware to support all those users and you should consider
using a hosting provider instead.

Similar Messages

Optimum Configuration for Parallels 5 and Windows 7 Home Premium 64-bit

I recently installed Parallels 5 and Windows 7 Home Premium 64-bit on my 15" MacBook Pro (recently purchased). What is the optimum configuration for the Virtual Machine?

Check Parallels forum and FAQ

Optimum Settings for Creating .MOV for Web

Hello,
I shot my video footage using a Canon XL-1 with a MiniDV tape. I then loaded this footage into Final Cut Pro (Version 4), running on MAC (Version 9.0), and then completed my editing. Besides creating a final 42-minute version to be shown on television, I would like to break up this 42-minute version into three separate pieces (a 21-minute, an 11-minute, and a 10-minute slice) for Web/Internet delivery.
To do this, I will create three separate .MOV files; however, I would like to know the 'OPTIMUM' configuration Final Cut Pro settings/considerations before creating these .MOVs for optimum audio/video quality when shown on a website. By the way, after I produce these .MOV files, I plan to encode them using FlixPro's ON2 V6 Codec to finally produce Flash 8 .FLV files for uploading to my web server.
Any and all advice regarding optimum Final Cut pro settings/considerations before creating these .MOVs would be aprreciated immensely!

Thank you for your reply. I take it then that current default settings are OK.

9i Configuration

We have recently moved our database to a SANS drive on an IBM blade server. This machine has 4GB of memory. However, no one has bothered to reconfigure the database. The current settings are;
Shared Pool 96MB
Buffer Cache 48MB
Large Pool 48MB
Java Pool 64MB
SGA_MAX 273.070MB
PGA Target 24MB (I have already increased this to 100MB)
My question is what should these values be?
I am familiar with MSSQL Server and have been thrown in at the 'deep end' here. Plenty of texts tell me how to change these parameters but I haven't found any yet that help me calculate what our optimum configuration should be.

hi,
looking at the PGA I use the following rule of thumb to start with
The first question we will have when we set this parameter is what is the best
value for it ?
- For OLTP systems
PGA_AGGREGATE_TARGET = (<Total Physical Memory > * 80%) * 20%
- For DSS systems
PGA_AGGREGATE_TARGET = (<Total Physical Memory > * 80%) * 50%
then there are some views that you can look at to tune further
SELECT round(PGA_TARGET_FOR_ESTIMATE/1024/1024) target_mb,
ESTD_PGA_CACHE_HIT_PERCENTAGE cache_hit_perc,
ESTD_OVERALLOC_COUNT
FROM v$pga_target_advice;
The output of this query might look like the following:
TARGET_MB CACHE_HIT_PERC ESTD_OVERALLOC_COUNT
63 23 367
125 24 30
250 30 3
375 39 0
you need to set your PGA so that there is no over allocation.
rgs
Alan

P45 Platinum Optimum memory settings with 2GB Corsair 1066 memory

Hi all,
Looking for some advice on optimum memory settings as I seem to have been getting unstable results from experimenting with settings and the optimum defaults.
My configuration is as follows:
MSI P45 Platinum (V1.0 PCB) (V1.4 Bios)
Intel Dual Core Duo E8500
Corsair TWIN2X2048-8500C5D DDR2-1066 5-5-5-15-2T configured for 2.2v in Bios
Palit 9800GT
2 x Maxtor 1TB HD's
1 x Floppy + Memory card readers
2 x Samsung DL-DVD-DW's
Silverpower SP-400P2C 400W PSU
With the default optimised settings loaded, the system does not book reliably or pass the memory test, in fact memtest86 just hangs when booted
With the Memory configuration set to 1:1.33 (1600:1066) the systems appears to pass the memory tests and windows is relatively stable. Voltage set to 2.2volt with Spead Spectrum disabled.
Is this really the optimum configuration ? as the Adjusted DRAM Frequency reported by the bios is 887 and I was expecting to see 1066.
Any suggestions as how to optimise reliably would be appreciated.
Although I suspect I may have a PSU issue as once booted the system locks when running any 3D games/demo's and it looks liek the 9800GT's need a lot more amps that I thought when putting the system together.
Rgds

Quote from: sylvester on 11-October-08, 20:43:36
Intel Dual Core Duo E8500
Corsair TWIN2X2048-8500C5D DDR2-1066 5-5-5-15-2T configured for 2.2v in Bios
With the Memory configuration set to 1:1.33 (1600:1066) the systems appears to pass the memory tests and windows is relatively stable. Voltage set to 2.2volt with Spead Spectrum disabled.
Is this really the optimum configuration ? as the Adjusted DRAM Frequency reported by the bios is 887 and I was expecting to see 1066.
First of all, your DDR2 modules are in fact DDR2-800 (they're using overclocked-overvolted DDR2-800 chips).
Sencondly you need to understand how a FSB/DRAM divider works.
Your processor is Intel E8500 and that means a bus frequency of 1333MHz (333MHz quad-pumped).
In your case 1:1.33 means 1.33 x 333 x 2 = 887 MHz for the DDR2 modules.
Set the divider to 1:1.6 if you want the DDR2 modules to run at 1066MHz.
And finally, the "optimum configuration" for max system stability means real DDR2-1066 (7-7-7-20 at 1.8V) or
DDR2-800 (5-5-5-15 at 1.8V).
cheers

Photoshop CC is dramatically slow when resizing or drawing.

Hello everyone.
Please assist me in finding out a bottleneck:
OS: Windows 7 x64 Professional
Photoshop CC, not 2014.
Hardware configuration:
MotherBoard- Asus P8Z77-V LX LGA
CPU - Intel Core i7 2700K + Zalman CNPS10X EXTREME
Video Card- Asus GTX660-DC2OCPH-2GD5, 2GB, GDDR5
RAM: 2x8Gb HYNIX Original DDR3 1600, DIMM + 2x8Gb Kingston KVR16N11/8 DDR3 1600, DIMM
HDD:
4xSeagate Barracuda ST2000DM001 - 2Tb - RAID 5 - for Data storage, OS TEMP directory.
1xWestern Digital 500Gb - for Photoshop scratch file, 300Gb free.
1xWestern Digital 2Tb for other stuff + OS page file. 1.2Tb free.
SSD: OCZ Agility 3, 90Gb - OS + Photoshop installed. 30Gb free
PSU: HIPRO HPP600W-80Plus
Tablet: Wacom Intuos Pro M.
Photoshop preferences:
History levels - 15
Cache levels - 8
Cache tile size - 1024Kb
Photoshop can use up to 70% memory.
GPU Acceleration - tried both "off" and "basic"
Previews for layers are set to "Minimum".
My main photoshop usage is in editing photos - 5760x3840 pixels.
When I create a document with 5760x3840 pixels and add several layers with editing, curves, masks and so on I try to resize the image. Photoshop starts lagging. Same effect happens when I create a mask with a simple brush. As example - I draw a 1cm line on my tablet, and then PS does the same with an about 10 seconds delay.
Any help would be appreciated. I had no same issues with CS6.

1. You probably have a specific problem with drivers that's causing your tablet operations to be very inefficient. I would look into getting new Wacom drivers to start with. Focus first on why your powerful system is running poorly with the tablet. Determine whether a mouse works better than the tablet, to ensure you've isolated where the problem is.
Then, after working that out, if performance is still troublesome to you...
2. I would suggest that, assuming you have the budget, you relegate ALL your spinning hard drives to be backup or low access storage drives. Do NOTHING with them for normal, everyday use. See item 3 below for that.
3. Get 2 TB of SSD storage in the form of four 512 GB SSDs, put them in a RAID 0 array - not RAID 5. Make that drive C: and install Windows, all apps, your working data, scratch, swap, etc. Don't worry about reliability - get good ones, pro models. They have high MTBF, and even with 4 SSDs you will not be taking nearly as big a chance of failure as with a single HDD. Re-evaluate whether the RAID controller you're using is a good match for high throughput SSD storage and get a good one if not.
4. Institute a thorough backup regimen. That sounds a bit scary, but it really boils down to setting your system up to do system regular system image backups to your HDD storage (or an offline USB drive). RAID 5 wastes resources in an ongoing way in anticipation of failure. Backup prepares you for any eventuality, and RAID 0 uses the resources efficiently - especially SSDs.
5. You have 32 GB of RAM, which is sufficient, but make sure you've done whatever is needed to ensure your system maximizes its efficiency in accessing that RAM. I'm not familiar with that ASUS motherboard, so I don't know what it can do, but sometimes you can dig and find info on optimum configurations where when the right things match, the system switches into a higher mode with more channels, etc. I know my workstation does special things when the RAM all matches.
I've personally done all the above, save for the tablet diagnosis as I don't have a tablet. There is NOTHING that will wake up a system - every aspect - like having a high throughput I/O subsystem. You'll wonder how you ever could stand using HDDs.
-Noel

A problem with channel strip outputs in Logic Pro 8

I have a Power Mac G5 2.3GHz dual core processor with 2GB memory, running LogicPro 8 connected to a MOTU 2408 Mk3. I originally assigned the outputs on the individual channel strips on my Logic default work page to the individual outputs on the MOTU, (1, 2, 3, 4, etc.,) as this is the optimum configuration for the amount of channels required for the recording I do. Checking theses settings recently though, I find the menu only gives the option to set outputs to 1-2, 3-4, 5-6 & 7-8? What has happened and how can I change them back?
Thanks

Sounds like your channel strips are stereo and not mono.
Click on the little linked circle button (format button) in the bottom of the channel strip to set it's mode to MONO (single circle).
If you have effect plugins inserted on the channel strip make sure they are mono and not stereo. (Click and hold on the effect on the insert slot and choose the mono instance of the plugin from the plugin menu).
Hope that helps
R

SWFLoader Problems

I have a major application that involves a Flex application acting as a main container, and it loads loads other Flex applications into it via SWFLoader.
Originally, the main container was built in Flex 3 and I was loading Flex 4 child applications into it. This was working fine until the Flash Builder Beta 2 came out. So I decided I am going to port my main container from Flex 3 to the Flash Builder Beta 2 and I am having all sorts of problems with the SWFLoader implementation.
I want to have two different configurations for the SWF content to load into the SWFLoader.
The first configuration should be totally sandboxed such that the only way that the child SWF can communicate with any other SWF is via LocalConnection.
The second configuration should be such that individual child SWFs are in sibling ApplicationDomains so they can have their own Singletons of the same name, but, content needs to be able to be dragged and dropped from one child SWF to another child SWF. For instance, the child SWFs will often have AdvancedDataGrid implementations and they'll often need to be able to have content dragged from one to the other.
I currently have a SWFLoaderFactory class that has a static method as such:
        public static function getSWFLoader(sandbox:Boolean=true):SWFLoader
            // create the SWFLoader instance
            var swfLoader:SWFLoader = new SWFLoader();
            // not sandboxed
            if(!sandbox)
                swfLoader.loadForCompatibility = true;
            // sandboxed
            else
                var context:LoaderContext = new LoaderContext();
                context.securityDomain = SecurityDomain.currentDomain;
                context.applicationDomain = new ApplicationDomain();
                swfLoader.loaderContext = context;
            // return the SWFLoader instance
            return swfLoader;
This was working exactly as I needed it to when the main container was in Flex 3 but now in Flex 4 it is broken.
What happens is with the swfLoader.loadForCompatbility set to true, the following error occurs:
TypeError: Error #1009: Cannot access a property or method of a null object reference.
    at mx.controls::SWFLoader/initSystemManagerCompleteEventHandler()[E:\dev\gumbo_beta2\framewo rks\projects\framework\src\mx\controls\SWFLoader.as:2466]
    at flash.events::EventDispatcher/dispatchEventFunction()
    at flash.events::EventDispatcher/dispatchEvent()
    at mx.managers::SystemManager/initHandler()[E:\dev\beta1\frameworks\projects\framework\src\m x\managers\SystemManager.as:3197]
I'll add that the child SWFs will always be on a different domain than the parent SWF container.
I would appreciate it if someone could help me solve this problem once and for all and create the optimum configuration that will allow me to do what I want to do.

If it crashes consistently, file a bug with the SWFs involved.

Need help solution to continually growing files

Have 3T and 1T external drives, an imac 500G drive and want to figure out an optimum configuration for evergrowing backup/itune/photo files. Previously cleaned out unnecessary/unwanted files. Would like to set this up to last a couple of years.
I am 61 and an APPLE novice (can hear me whimpering??). Suggestions would be GREATLY appreciated. Thanks

See my FAQ*: http://www.macmaps.com/diskfull.html
Consider having an authorized service technician install a larger internal drive, or burn your data off to DVD.

HSRP For the Gateway Redundancy.

Hi all
i just need a simple how to configure 2 Routers(R1;R2) to run HSRP For the Gateway Redundancy ,if one of the 2 routers Fail.should i connect the 2 routers 2gather via cross cable.than one straight cable to the 2 separate distribution switch.(2 Etherchanel configured between Dist switch)PS LIST ur optimum Configuration
Ur help very much Appreciated

Hi,
i think you talking about campus network where you have two distribution two access and two core router's.
With that prospects.
My suggesion will be to have etherchannel between distribution switches and both the distribution switch should be connected to both the core router.
than use HSRP in distribution swithes.
configuration and diagram is given below.
do let us know if you want any more information.
interface FastEthernet2
ip address 172.69.90.1 255.255.255.0
standby priority 200
standby preempt
standby ip 172.69.90.6
interface FastEthernet3
ip address 172.69.91.1 255.255.255.0
standby priority 200
standby preempt
standby ip 172.69.91.6
like the above configuration you can configure second switch also you can apply on vlan interface too.
HTH

11501 Web Mgmt Locks Up

Hello,
I've got two 11501's running 7.50.0.04 that locks up at variable points while accessing the Web Management interface (HTTPS). Most times it locks up and requires a CSS reboot during login, but we have been able to get into the various menus, and then experience our lockup.
Fortunately, these switches are not in use - yet, but will be very shortly. We would like to use the management interface for various report retrieval. We also notice the CSS will not respond to CLI telent session during these lockups.
Is this a problem with the CSS? I cannot find anything in the CSS's logs that might help.
I'm a newbie to these things and with the vacuum of publications it's really difficult to grasp all their possibilities and optimum configurations. All help is appreciated.
Thank You,
Jim

Jim,
is the console still accessible at the time you can't telnet or when your Web session hangs ?
are you accessing the CSS via the management interface or another port ?
This issue you describe was reported already but we were unable to capture enough information and the problem disappeared by itself.
CSCei15542
Accessing CSS GUI via the management port causes hang
Can you use another port ?
Are you willing to troubleshoot the issue and capture the info we may need ?
Thanks,
Gilles.

When I sync ical on iphone with mac, every entry is multiplied. What are the settings for syncing without repeats?

When I sync ical on iphone with mac, every entry is multiplied. What are the settings for syncing without repeats?

If your wireless environment is less than ideal, as it is in most inhabited areas, the influence of other wireless networks and devices competing for the same very small portion of the electromagnetic spectrum will affect your network's performance. Your own wireless network is a microcosm of that environment.
To mitigate the effects of competing wireless networks, use 5 GHz to the maximum practicable extent. It is comparatively free from interference contributed by other wireless devices and is likely to remain that way for some time. However, it is less tolerant of intervening obstructions and construction materials than the 2.4 GHz spectrum, so it's not a panacea. Every installation is different so you must experiment to determine the optimum configuration.
Place your wireless router (what kind is it?) in a central location, convenient to all the devices that require it. Most of the time a utility or ISP will locate it in some remote corner next to the utility service entrance. That is almost always the worst possible location.
Others find that only a direct wired Ethernet connection between router and devices will maintain uninterrupted access to Airplay.
That is almost certain to result in the best possible AirPlay performance. Short of having everything hard wired (it's not an option for the iPad for example) use an Ethernet connection for all the devices that lend themselves to that method. Your AppleTV is certainly one of them, since its location is not likely to change very often. Your AirPort Express should be another, since your audio systems are likely to stay put as well. Connect them both to your existing router, either directly or through an inexpensive Ethernet switch. Your problems are likely to be solved with just those two connections.

Streaming iTunes to my Hi-Fi

I'm trying to determine the best way to stream music from one of my Macs to an aged analogue British Quad HiFi system (with phono/rca inputs).
I do have an Airport Express which I assume I could use and I have a lead to take the signal into the Quad via the RCA inputs. My problem & question is, "is this the right route?" because my wireless router is some 20+ feet away from the hifi system. The router serves 3 Macs wirelessly, and as I understand it, that I would need to connect an appropriate lead from the router to the Airport Express which is too far away. ( I would need to connect the Airport Express to the Quad system via its three foot long lead).
So I'm rather back to the drawing board at the moment. I did consider buying the Airport TV, as my HDMI Sony TV is only 9 feet away from the hifi. But I'm unsure what cabling I would need to purchase. If, in fact, this was practical.
It's my intention to part company with my 300+ CD collection as this is now all stored in iTunes. So I would appreciate some help please in pointing me in a direction that is feasible. Unfortunately, the hifi and my main Mac at at opposite ends on my home which doesn't exactly help!
Thank you.
Barry

Barry Wilbraham1 wrote:
... I do have an Airport Express which I assume I could use and I have a lead to take the signal into the Quad via the RCA inputs. My problem & question is, "is this the right route?"
That is how the Express and AirPlay streaming is designed, so it will work. The most optimum configuration minimizes the number and distance of wireless links in the path from the Mac, to your router, to the Express.
20 feet is not very far so the wireless connection from the Express to the router ought not to be a problem. The distance from your Mac to the router, which you say are located on opposite ends of your home, could be.
The audio cable length from the Express to your Hi-Fi is not limited to 3 feet. The one you have may be 3 feet but there is no reason you can't use a much longer one. I doubt the 20 foot distance will be a limiting factor though.
The best way, as you originally asked, is to connect the Express to your router with an Ethernet cable, assuming your router has an unoccupied Ethernet port. Even better would be to connect your Mac to the router the same way. That would eliminate all uncertainty. Those cables can be up to 100 m long. Having such an all-wired network may be impractical though.
It is impossible to determine how well wireless streaming will work due to the number of variables involved in any wireless installation.

UC540 with Allied Telesis 8000S Switches

Hi All,
This is my first post - so please be gentle!
One of the sites I look after uses Allied Telesis switches and routers (and this is what I trained on) with a UC540. I see that the UC540 is attached to the AT8000S switches via the expansion port. Can someone recommend the optimum configuration in this scenario? Should the UC540 be trunking through one of the switch ports or the expansion port? I know that the UC540 can configure the Catalyst switches via the expansion port so should the AT switch be connected here? Also, since they are using non-Cisco switches I think I should enable LLDP-MED and disable CDP - what are your thoughts?
Regards, Shayne Harris - New Zealand

Hi Paolo,
Thanks for your input.
I will have to make do with the AT-8000S swtiches as we won't change brands - so I need to make this work.
From what I have noted I need to turn off CDP and turn on LLDP-MED. I also need to ensure the FX port is set to Trunk and Vlan1/Vlan100-Tagged. Is the EXP port ethernet 0/1/8 ? I think I should also enable spanning tree port-fast on the switch also. Currently, the UC540 is sitting on a totally different wired infrastructure, the customer is moving to a new building and I want to take the opportunity to re-configure the UC.
Thanks for your help
Shayne

CUDA in Matlab with MBP

I'd like to run CUDA (via MatLab) on my Macbook Pro. With two graphics cards, it seems I should be able to run the screen on the integrated card, and CUDA on the discrete card- basically to maximize the gpu memory and computational power for CUDA. Is there a way to set this up?

Interesting....right now I'm just experimenting. I'll have a look at the Jacket at some point and see how it compares- the sparse linear algebra would probably help me a lot, which apparently PCT doesn't do. The problem I am primarily trying to solve is an mldivide for a large, sparse matrix. With PCT, I've just been working with smaller matrices at the moment, since it doesn't take the sparse allocation. For now, I just want to get something running, though, and was trying to get my system in an optimum configuration first. The other problem with Jacket is I'm in academia, where the economics work a little strangely. We won't sell our software, but rather give it away in hopes of people citing our papers, which then get us grants. Which means that adding a somewhat expensive library requirement will mean people may hesitate to try it in the first place....but we'll see. MatLab has a bit of a death grip on academics.
I've actually achieved what I wanted- using the integrated card for the graphics and the discrete card for computation. I used the energy saver preferences to turn on the discrete card, and gfxCardStatus to make the system use the integrated card. It doesn't work every time, though, and is horribly unstable. But, that makes me think that there ought to be a more stable way to do it.

Optimum Configuration

Similar Messages

Maybe you are looking for