Latency increase

Joined BT recently and for first 5 weeks or so been fine but this last week my Ping has increased from about 30 ms to + 50 which is killing the xbox/ps3 ,sometimes as high as +100ms
Microsoft Windows [Version 6.1.7600]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.
C:\Users\MEDION>ping www.google.co.uk
Pinging www.l.google.com [209.85.143.104] with 32 bytes of data:
Reply from 209.85.143.104: bytes=32 time=53ms TTL=50
Reply from 209.85.143.104: bytes=32 time=53ms TTL=50
Reply from 209.85.143.104: bytes=32 time=51ms TTL=50
Reply from 209.85.143.104: bytes=32 time=50ms TTL=50
Ping statistics for 209.85.143.104:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 50ms, Maximum = 53ms, Average = 51ms
Line state
Connected
Connection time
1 day, 6:33:14
Downstream
8,128 Kbps
Upstream
448 Kbps
ADSL settings
VPI/VCI
0/38
Type
PPPoA
Modulation
ITU-T G.992.1
Latency type
Fast
Noise margin (Down/Up)
8.4 dB / 23.0 dB
Line attenuation (Down/Up)
31.0 dB / 16.5 dB
Output power (Down/Up)
19.8 dBm / 12.1 dBm
Loss of Framing (Local)
0
Loss of Signal (Local)
0
Loss of Power (Local)
0
FEC Errors (Down/Up)
0 / 0
CRC Errors (Down/Up)
350 / 2147480000
HEC Errors (Down/Up)
nil / 0
Error Seconds (Local)
1140

Low latency cannot be guaranteed anytime, especially by BT, whose policy for xdsl is for stability over speed. Depending on which RAS you're connected to, and network traffic, it's a lottery. Running fastpath has seen your margin increase, and an increase even by 3db reduces signal by 50%, as it is logarithmic. You also have CRCs, resends, uncorrected dumped packets, demanded by the router ... because of fastpath ... that slows transmission down. Even running deep interleave would only add about 30 m/secs to transmission, you're nowhere near that level of interleave. Interleave may benefit you, your choice.
Click the star if my posts have helped.

Similar Messages

Latency increase with large number of databases

Hello,
I'm testing how BDB performs on random updates as the number of databases increases. I'm seeing increase in latency as the number of databases goes up. and strace shows that more time is spent on futex() when there are more databases. Is this an expected behavior?
Thanks!
--Michi
Set Up:
<li> BDB version: db-5.1.19
<li> Operating system: RHEL4
<li> Number of records: 2.5M
<li> Record key size: 32B
<li> Record value size: 4KB
<li> Page size: 32KB
<li> Access method: BTREE
<li> Records are inserted into a database based on the hash of the key.
<li> BDB is accessed via RPC server
<li> RPC server thread pool size: 64. All the threads share the same env and db handles.
<li> Number of RPC client process: 64
<li> Each client does random updates. Throughput is throttled to be about 250 requests/sec, which means each process does about 4 requests/sec.
Flags used for env and db:
ENV flag DB_THREAD | DB_RECOVER | DB_CREATE | DB_READ_COMMITTED | DB_INIT_TXN |
         DB_INIT_LOCK | DB_INIT_LOG | DB_INIT_MPOOL
DB flag DB_AUTO_COMMIT | DB_CREATE | DB_THREADDB_CONFIG
mutex_set_max 2000000
set_cachesize 1 0 1
set_lk_max_objects 2000
set_lk_max_locks 10000
set_lk_max_lockers 10000
set_lg_regionmax 8388608
set_lk_detect DB_LOCK_DEFAULT
set_thread_count 20000
set_lg_max 1073741824 Result
Number of databases: 256
Average latency: 52 milliseconds
strace output
% time     seconds usecs/call     calls    errors syscall
64.74 121.199645        1032    117398           recvfrom
17.21   32.217921         487     66104      9335 futex
8.76   16.409370        2724      6024           pread
2.47    4.630256         813      5694           fdatasync
2.19    4.102371         105     39162           sendto
Number of databases: 512
Average latency: 123 milliseconds
strace output
% time     seconds usecs/call     calls    errors syscall
58.33 121.384230         933    130158     20045 futex
23.06   47.977408         383    125146           recvfrom
4.41    9.167276        1425      6431           pread
3.11    6.472676         155     41716           sendto
3.08    6.414824         366     17506           sched_yield
Number of databases: 1024
Average latency: 133 milliseconds
strace output
% time     seconds usecs/call     calls    errors syscall
70.46 102.981390        1077     95656     14884 futex
9.88   14.435594         169     85256           recvfrom
5.48    8.008787        1825      4389           pread
3.13    4.576880         368     12436           sched_yield
2.98    4.351077         153     28415           sendto
...

Hi James,
Presently, we never evict a DB after it's opened or encountered during recovery. Each DB takes about 2,000 bytes. So if you have 16K DBs you need approximately 32MB of memory, assuming all of them could be opened or recovered during a process lifetime. Unfortunately, even closing them does not cause them to be evicted. If you encounter them during recovery, this will also pull them into memory.
We have an FAQ entry on this:
http://www.oracle.com/technology/products/berkeley-db/faq/je_faq.html#37
So you will need a bigger cache size. If you are encountering this during recovery, then you could try a more frequent checkpoint interval.
I hope this is useful.
Regards,
Charles Lamb

Latency Increased or Reduced by Multicore Support? MS2 vs. Kore 2?

I'm putting together a live software synth rig that I'll be using alongside a few cherished hardware pieces. I'm a current Digital Performer user but I really need a different host for live gigs. I'm going to be running Native Instruments' Komplete 6 a lot. I'm really torn between buying Logic and Main Stage or Native Instruments' Kore 2.
They each come with a comparable assortment of great instruments and effects but definitely seem to have their own strengths and weaknesses.
I'm pretty hesitant to buy a copy of Kore 2 right now because NI really hasn't gotten Komplete 6 integrated with it yet. On the other hand, it'll be a powerful librarian for my NI software and ultimately add a significant degree of flexibility these synths as well. That being said, I'm seeing all sorts of people struggling to find workarounds for and odd solutions to what I consider to be relatively basic issues from a hardware synthesis standpoint.
Main Stage 2 definitely looks like a great program on paper. It's a lot less complicated than Kore 2 is, but at the end of the day, I don't want to invest an insane amount of time in programming my live VST/AU host. I want to be able to get my sounds on stage quickly and efficiently, I want to be able to recall them instantly when I need them in the set, and I want to avoid instance redundancies or other obnoxiously RAM-consuming situations (MIDI program change recall).
I'm running a Macbook Pro (2.4 ghz Intel Duo Core, 4 GB RAM) and I know that Main Stage 2 has multicore support while Kore 2 does not. Intuitively, one would assume that using half of your CPU processing capabilities would be inferior. However, if you're using two processors to accomplish one task, you're theoretically going to have more latency because they have to communicate with each other. So Main Stage 2's multicore support may actually prove to be a hindrance on stage.
I'm definitely very serious about not spending an eternity programming my set list into a new host, but if I had to choose between a simpler platform or lower latency, I'd undoubtedly side with low latency as it's absolutely essential to the music I play. After years of practice I'm a player first and foremost, songwriter second, synthesist third, and gear-nut fourth. I don't have the RME Fireface 800 which I know will offer the best latency of any Firewire interface, but I'd like to see if my Presonus will be usable before I invest that heavily in a first-generation FW800 unit.
So I've got a few questions for you guys...
1) What kind of latency figures are you seeing on stage with MS2? What are your laptop specs and what interface are you using? Can any of you get it down to ~5ms?
2) Have you had any experience with Kore 2 or a single-core live VST/AU host?
3) Did any of you buy Logic Studio primarily for Main Stage 2? What informed this decision?

I am getting under 4 ms with a 1st gen. MOTU Ultralite on a 2.2 Ghz MBPro.
The MOTU units are well-known for low latency. This is much better than any other interface I have used, but I have not used the FW800.
Your mileage may vary.
Thomjinx

Why has my Latency (Ping) increased from 20ms to 3...

Got BT Infinity installed (FINALLY, after over 12 months of them putting it off every 3 months) but this was due to a problem in our village I believe. Anyway, I am right at the end of my line and 5KM or so from my local exchange, I was only getting 2.3 meg max on ADSL and 0.8 meg upload, on BeThere (02) broadband. My pings averaged 18ms to BBC.co.uk (London/Local Servers).
Now, had infinity installed last Tuesday, and expected upload and download speeds according to Infinity of 42.1MB download and 6.5MB upload. Fastest speedtest yet was 48MB on speedtest.net and fastest upload was 6.7MB.
Now two questions:
1. A few days ago, I was receiving speeds on average of about 42-44MB in speedtests, now as of yesterday morning, this has dropped to around 38Meg. My Upload is about the same at 6.2meg.
2. THE MAIN QUESTION - Why has my Latency increased from about 20 (18-23ms on average) to 30?
Every speedtest I do now for the last 2 days, is coming back with a ping of 29-33ms which is a sudden increase of 10ms. I'd like to know what may have caused this, as I game and I know a better ping is crucial for gaming.
Can anyone shed some light here?
Thanks

Nick134 wrote:
Well as you can see, I've attached you a screenshot of my tracert command results, what do you think?
Well from that tracert you can see the increase isn't all down to interleaving, the first hop after the router is only 15ms, that's the interleaving, it's after this hop that the ping times increase to near 30, so in your case the ping increase is all down to how bt are routing your connection over their network.
Here's my tracert to the same website, routing is different for me and hence the better pings, also the lower interleaving depth on the second hop is also helping of course

Windows Server 2012 - Disk IO & High Latency

We have a very strange issue on what appears to be all of our Windows 2012 servers.
We can build a basic server, HP Proliant DL 380 Gen 8 - RAID 0 (2x 300GB 15K SAS). Install Windows 2012, patch to latest levels, all HP firmware and drivers at latest levels on HP website.
Pretty much a vanilla Windows build, even without Anti Virus at this stage. (This same issue occurs with IBM HS22 blades as well with base build and all latest IBM firmware / drivers so does not appear to be single vendor issue).
Take a large file, for testing a 20GB VHD. If we copy this file from C:\Temp to C:\Temp we get about 1.4GB/s transfer speed (Not bad). However post copy all processes in Performance Monitor
Disk Latency increase to over 1500 ms. This happens every time we copy a large file.
The file systems is NTFS.
The array controllers cache settings are set 75% READ / 25% WRITE.
This problem is much worse when it occurs on our Windows 2012 Scale Out File
Cluster with Hyper-V hosts using the storage. In theory if a client (Hyper-V server) requests a large Read or Write activity you see a good throughput, but then followed by a large period of disk latency afterwards on all processes on the
file server. If multiple requests from clients occur simultaneously then this problem is severe enough to hang / crash the file server.
This also occurs when the servers are not using the "C:" drive and instead are using the CSV volumes on our IBM V7000 array. In this scenario the server has 4x 8Gb HBAs connected to
the 8x 8GB FC connections on the V7000. EasyTier is enabled SSD Disk Pool to 15K SAS Disk Pool. So in this scenario I know that the issue is not due to an underlying performance issue within the disk subsystem.
This does disk latency issue did not occur on Windows 2008 with the same hardware EG: HP Proliant DL 380 G7 with local or V7000 disk.
Therefore there appears to be a major difference / problem in the Windows 2012 disk IO stack.
Another thing to notice, if a process is running a large copy EG: the VHD copy above and a second or third copy is executed the data transfer of all process does not reach an equal level instead
one process reaches the highest transfer rate then slows down and the other process takes over. This kind of activity is again not what we used to see on 2008 or other operating systems.
Can anyone shed any light on this? Does anyone else see a similar IO profile?
We have a case with Microsoft support for weeks with no resolution.

Hi,
We do have multiple SoFS nodes in the cluster, all have have dual 10Gb NICs with RDMA ability (Mellanox Connect3) connected to Brocade VDX switches. RDMA works a treat between hosts using SMB Direct & SMB Mulitchannel. The SAN disks from SoFS are 8Gb
FC, the local disks are SAS.
The problem we have is locally not even remote. EG: The latency occurs on the SoFS when copying data from C: to C: or from CSV Vol1 to CSV Vol1.
It is very strange, not seen this issue before. It makes me think its a bug as it can be created on other Windows 2012 hosts with just local disk.

Windows 7 and Networks with High Latency

We are currently trying to rollout Windows 7 on our network to replace XP but have encountered an issue whereby we have remote clients that access the network over high-latency satallite links (BGAN and Vocality/Satellite). The latency of the links (based
on ping results) can be 600ms for Vocality and 1-3s for BGANs.
The particular services that don't work are a full motion video solution using TVI Viewer 7.9.1 and Outlook 2003 or 2007. These work fine on Windows XP. Shares can be accessed but are significantly slower than XP and ping does respond fine.
Windows 7 is running on Panasonic Toughbook CF52 and CF74 and I've tested in with a vanilla install with no updates and not on the Domain (2008R2 native) to eliminate GPO interference and tried it with all MS updates as of about 2 months ago.
I've tried removing the extra services on the network card (Topology Discovery, ipv6 and QoS), updated to the latest NIC drivers from Panasonic and drivers from Intel themselves. Reduced the MTU to as low as 500 and increased the Frame size (I forget what
to but was following a guide for slow links).
I've successfully replicated the issue on out development system using a satallite simulator.
Windows 7 with Outlook and TVI work fine on our network when connected via the LAN, ADSL, 3G and WiFi.
I'm currently analysing Wireshark captures but they don't seem any different to the XP ones.
Any help would be much appreciated.

Hi,
I noticed that your issue just happened when you use satellite transmission connection.
The fact is that this kind of connection in Windows 7 use TCP protocol. Transmission Control Protocol ( TCP ) under ideal conditions can provide reliable data delivery, but it is inherent in the existence of a throughput bottleneck, with the emergence on
the long-distance WAN packet loss and latency increases, the bottleneck is becoming more prominent and serious. In satellite networks with high loss, effective throughput may be as low as 0.1% - 10% of available bandwidth.
However, FASP can be the solution.
FASP
http://asperasoft.com/technology/transport/fasp/#overview-464
This response contains a reference to a third party World Wide Web site. Microsoft is providing this information as a convenience to you. Microsoft does not control these sites and has not tested any software or information found on these sites; therefore,
Microsoft cannot make any representations regarding the quality, safety, or suitability of any software or information found there. There are inherent dangers in the use of any software found on the Internet, and Microsoft cautions you to make sure that you
completely understand the risk before retrieving any software from the Internet.
Thanks for your understanding.

Query latency vs number of partitions

I am trying to measure the average latency per object retrieval using the query approach, and what I am finding is that as the number of instances of the server (a server being a single jvm running DefaultCacheServer with backup-count set to 1, and partition-count set to 1021) increases, the latency increases significantly as shown in the results below:
Latency (microseconds) for Number of server instances
Objects 2 4 8 16
In Cache srvrs srvrs srvrs srvrs
1K 593 730 889 1175
10K 592 720 880 1206
50K 580 724 879 1200
100K 585 726 877 1204
The query code is as follows:
final Filter filter = new EqualsFilter(idExtractor, key);
final Filter limitedFilter = new LimitFilter(filter, 1);
final Set spotPricesKeys = cache.keySet(limitedFilter);
if (spotPricesKeys != null && spotPricesKeys.size() > 0){
spotPrice = (SpotPrice) cache.get(spotPricesKeys.iterator().next());
and
private static ValueExtractor idExtractor = new KeyExtractor("intValue");
Any idea why the latency is increasing with the added number of server instances?
Thanks
Adsen

Partitioning strategies is an "advanced" feature of Coherence. It works much like a hash function but instead of a hash code it calculates the partition number that will be used for each key. You can use more or less any deterministic algorithm based only on key information. Lets assume that you have a cache with strings as keys (representing for instance a persons name) and you would like to partition based on the first letter it could look something like this (this is just an example - in production code you would like to make sure the algorithm statistically makes as even use of all the partitions that you have in your cluster!):
public class MyPartitionStrategy implements KeyPartitioningStrategy, Serializable {
private PartitionedService partitionedService;
public MyPartitionStrategy() {
public void init(final PartitionedService partitionedService) {
this.partitionedService = partitionedService;
public int getKeyPartition(final Object key) {
if (key instanceof String) {
return ((String) key).charAt(0) % (partitionedService.getPartitionCount() + 1);
} else {
return key.hashCode() % (partitionedService.getPartitionCount() + 1);
With this filter you would know that all keys starting with the same first letter ends up in the same partition. You would specify the use of this filter once in the coherence configuration (dont remember the element to use on top of my head though).
Now assume that you for some reason want to do some query that only should apply to entries with a key starting with the letter 'a' you could do that with a PartitionedFilter (you would have to write some code that determines what partition the key 'a' would end up if it existed and then restrict the query to that single partiton). With this solution Coherence will only send the request to the single node holding the primary for that one partition holding all the "axxxx" keys and not to all the nodes in the cluster making the query scale close to linerary no matter how many nodes you have!
A problem when the number of partitions grow is to find a partitioning strategy that allocates a useful set of objects into a single partition. If you for instance have like 4000+ partitions and would like to divide addresses based on state in order to perfrom queries against all of your customers in a particular state there is no obvious way to do that since the number of states is much smaller than the number of partitions :-(
This kind of partitioning can however be very useful in many cases (we have used it to good effect in my current project).
In produktion code it is often some part of a composite key that is useful as a key partitioning criteria and you could then for instance return the standard hash code of that key part modulus he partition count.
Another possibly useful feature you may look at is when it comes to limiting the number of nodes invloved in queries is the KeyAssociatedFilter (if you use key association at all in your cache).
/Magnus

Built-In Core Audio latency wierdness...

Hi all,
Could someone verify this for me please. Using just the Built-In inputs and outputs on a MacBook Pro:
1. Connect something to the line input. Connect speakers or headphones to the MacBook Pro outputs.
2. Start Logic. (Choosing Built-in core audio in audio prefs if necessary, with Software monitoring enabled. Use low-ish buffer size eg 256 samples.)
3. Rec-Enable an audio track from line input.
4. Check Audio throughput is as expected i.e. negligible latency.
5. Unplug headphone/output.
6. Re-Plug headphone/output
7. Latency is now around 1 second!! What the...?
If anyone else can verify this I'd be most grateful. If it's just me (and I hope it is) then I can set about sorting it out myself with re-installs or whatever....
The strange thing is that audio instrument latency remains normal - it's just the audio input --> output chain that seems to be affected. Pressing play seems to temporarily reset the buffers again, although I am sometimes getting clicks, pops and jitters.
Cheers,
Bryan

Hi,
Of course, I realize that the Built-In in/out is not sufficient for professional audio work.
I'm really trying to get to the bottom of serious general Core Audio problems that I've been experiencing with all my audio devices: Motu828mkII, IK-Multimedia Stealth Plug (A Guitar cable) and Built-In. You may remember this thread about latency issues when recalling Channel Strips saved as Performances.
http://discussions.apple.com/thread.jspa?messageID=4903770&#4903770
Basically, core audio becomes totally unstable after recalling a channel strip saved as a Performance. Latency increases dramatically - until you press play again, which resets the latency, although core audio remains unstable i.e. glitches and other strangeness. Identical behaviour when plugging/unplugging headphone out.
I do hope I'm not whining too much, it's just that I'd like to be able to use MBP to process live guitar, and at the moment it is unstable.
Cheers,
Bryan.

Soundbanks percussion kit and EAX microphone problems

I bought the PCMCIA Audigy ZS notebook to use it with my notebook HP Pavillon ZS6000 (Athlon 3200, 1 Gb RAM) having the hobby to sing with friends using karaoke software like Van Basco or Karawin.
I bought it in a blister pack at the end of August.
In both sw (Karawin or Vanbasco) you can set the output device to one between internal midi map, Microsoft GS wavetable, Creative soundfont player, or a virtual software like YamahaXG or Virtual Sound Canvas if you have.
I use the Creative soundfont player and change the banks through soundbank manager. It works properly, but if I try to change the percussion kit, dont' hear any difference, like if it use only the first one. Moreover, if I try to combinate 2 soundbanks loading over a second one only with a great piano for example it sounds only the first bank in the list.
Another problem is with Eax anvironment, that produce on the microphone sounds only a continuous bipped sound (I dont' know if it's the one you called bopping) and no echo or reverb...
Least if I use the internal speaker the latency increase terribly to more than 1 second.
Someone can help me?

$? Hi.
I have got same problem here.
When I switch from "creation to game mode" or from "game to creation mode" I can hear microphone/line in playback.
It is muted on the panel but I can HEAR it.
I always have to switch it ON and OFF (in both modes).
Same thing happens when I play through ASIO.
Microphone playback activates but panel visualisation is OK.
This behaviour is pretty annoying.
How is THIS even possible ? I can not believe it.
One more thing about create console I do not like.
It is SLOW !
Slow loading (opening), slow switching.
And two words to Creative developers:? "SIMPLICITY" and "EFFECTIVENESS".
driver 2.7.0007
vista x64
4GB
G3+pent E6300
listening through digital out

Can't upgrade Mainstage 2.0.1 to 2.1

Hello,
I've tried to upgrade Mainstage 2.0.1 to 2.1 but each time I do I get a message that its not necessary for my computer. But, Version# is still 2.0.1
Logic Pro upgraded to 9.1.0 no problem.
Whats up?
Thanks for any help!
Running on:
MacBook Pro 15" 4G RAM OSX 10.6.2

No I didn't rename the application.
I finally called Apple support for some help. And help I got! The tech was patient, knowledgeable, and professional. He talked me through some system, library, and terminal inquiries and decided that I needed to remove both apps, no other associated, Logic Pro 9 and Manstage 2, and reinstall from DVD disk, then install both upgrades Logic 9.1 and MS 2.1. Everything works fine so far! Well sorta..
I did noticed in MS a substantial latency increase that I didn't have in MS 2.0.1. I made some memory buffer changes in the MS preference/audio to 512 which was a little more acceptable.
MS does seem to be a little more sluggish sometimes when making edits of patches.. so I'm a little concerned about that.
Any suggestions about reducing latency and general MS sluggishness to this upgrade would be appreciated.
Thanks again for your ideas.

Mac Pro RAID 0 setup with bootcamp

I'm about to take the plunge and buy a Mac Pro as soon as I get the money (or credit) to. There have been things here and there that have turned me off but I found ways around some concerns, but this one I haven't been able to find.
I am a musician and want to utilize raw power that Mac does so well with. But unfortunately my DAW of choice doesn't support OS X. I hate Vista, I never upgraded, I stayed with XP but have been willing to change to OS X with XP as well, to get the best of both worlds. Then i find my DAW doesn't support XP 64 bit so i am going to have to use Vista 64-bit (bummer).
What I haven't been able to figure out is if it is possible to run a RAID 0 of the windows OS and all its data. The Mac Pro RAID card doesn't support any Windows OS so i am wondering if there is anyway possible to get the Windows OS in RAID, any good hardware or software that will accomplish this. Like if i get the Mac Pro RAID cad, is there a good software or hardware RAID that will allow me to install windows in a RAID 0 configuration?
If its possible I would have no reasons to not move over, but I really need to have my RAID setup in windows since I am using latency sensitive disk streaming programs that need to stream files in the GBs
Also could I just opt out for the Mac Pro Card and get a generic RAID card that supports both Windows and OS X? And if so any recommendations?
Message was edited by: jmoss211

I'm glad you are in the planning stages!
The question I didn't know to ask is always the one that bites the worst.
One of the best options might be the RR 4320
http://www.hptmac.com/US/product.php?_index=50&viewtype=details
http://www.barefeats.com/hard109.html
http://eshop.macsales.com/item/Highpoint%20Technologies/RRAID4320/
One of the first Mac Pro RAID and Boot Camp work?
Games, hence the whole gaming graphics area, has and is not high on the list of features, which had been a large reason for people installing Windows on Mac.
I didn't want to see you invest in SAS and not get performance level you want.
StorageReview Performance SCSI/SAS vs VelociRaptor
Here is the card Barefeats felt was the card that Apple has to beat, and points to some of the limitations of SAS on the Mac Pro. http://www.barefeats.com/hard104.html
MacWorld SanFran in a couple weeks is when Apple traditionally would use to announce some of the new products for the next year (Snow, Mac Pro '09, new iMac, etc).
Apple is not very upgradeable. It would cost me more to upgrade the cpu than buy new (though can get a good $$ on my system - in part because it runs the prior version of OS X 10.4 "Tiger". New systems will only run the current/latest OS, you won't be able to run Leopard if a system comes with Snow. The graphic choices and upgrades can be lame. It is almost a closed box.)
I've talked to people that were into audio that used Fibre Channel in part due to low latency, increased queue depth (which is one area SATA has never implemented, 32 vs 256 I/O per sec).
+Forget trying to use the Apple search to find threads, it is 'busted' if I can't find the thread mbean and I had with someone only last week on RAID5/6. I turned to Google to find this thread +
Multi-core compiler optimization is big on Santa Intel's wish list to get into the hands of developers and vendors.
*3 GHz Nehalem outperforms the latest Opteron by a margin as high as 80% and more.*
Intel has apparently allowed HP and Fujitsu-Siemens to break the NDA on the Xeon 5570 processor for PR reasons as both companies have published SAP numbers on a Dual Xeon 5570. The Xeon 5570 is based on the same architecture as the Core i7. It is a 2.93 GHz quad-core CPU with 4 times a 256 KB L2-cache and one huge shared 8 MB L3.
http://anandtech.com/weblog/showpost.aspx?i=532
And AMD just gave Intel a new run for the money.
AnandTech’s Johan de Gelas has taken a look at AMD and Intel quad-socket servers. The quad-socket six-core Intel Xeon X7460 at 2.66 GHz is compared to the quad-socket four-core Shanghai-based Opteron 8384 at 2.7 GHz. And what were the results of a test using VMware’s ESX Server 3.5?
Shanghai’s 16 cores outperformed Dunnington’s 24 cores (or 48 if Hyper-Threading was used) by 6.5% or 8.8% depending on whether or not IBM or Dell’s chipset was used. But if we normalize that out per core, Shanghai outperforms by 59.8% (per core) or 63.2%. That is unbelievable!
http://www.geek.com/articles/chips/amd-cleans-up-in-high-end-server-virtualizati on-20081223/

Bad Connection to French Servers

Hello, I am pretty satisfied with my BT internet normally, however, there seem to be a problem when i connect (play) on french servers, on any game. I did some tests with Windows CMD, i traced the ip.
http://i.imgur.com/VNoLtPn.png
I am not really good at looking at those ip address and 'ms' using the cmd on windows.
It doesn't appear to be at my end because one of my friends uses BT as well, pretty far from me in the UK and he has the problem.
Any ideas?
Cheers

That's why I asked for the pathping, we obviously interpreted two different problems looking at the picture. I wouldn't class a 100ms increase in latency a problem if packets are not getting dropped. It will mostly not have adverse effects on software. On the other hand, there's 3 packets lost on the picture, but on hop 10 not 11.
The latency issue is gone by now, btw, I can use the same route to ping the same servers with no 100ms latency increase. I can't test for packet loss because I can't perform a reference test where there isn't any.
Edit: actually, never mind, my hop 9 is a different btcore router. There's no 100ms increase on my route, but I'm still working on finding out which LINX router I'm on because it may be a different one. The route from the first french IP is the same.

Digital I/O Problems

I have a FlexMotion 7334 board with NI-Motion 6.0.1 and I am experiencing some digital i/o problems. After running my system in a loop for several hours, I unexplicably start having problems where my digital i/o starts functioning incorrectly (it does not detect something that is there). I was curious to know if there are any known issues with the digital i/o like will it freeze if the board is accessed too frequently or any issues like that. Please help. Thanx.
--firoz

Firoz, the DIO lines on the 7334 are static, not latching, so as you probably know a line must be constantly polled when you are trying to detect the change in the state of a TTL signal. This polling is software-timed and therefore influenced by software latency, increases in the PCI bus traffic, etc. Therefore if the duration of a pulse is too short, or if the DIO lines are not polled fast enough, you might miss a signal. Is it possible that this is part of the problem you are seeing?

Anchor mobility configuration getting lost in wlc 5508 ios code 7.4.100.0

It is observed that in WLC 5508 , ios 7.4.100.0 , mobility anchor configuration on wlan is getting lost . we configure anchor ip address on guest wlan > mobility anchor > Switch IP Address (Anchor).
We have configured the template on NCS 2.0 to push the anchor mobility ip address on all WLC
Has anyone oberved this behavoiur. We have more than 100 WLC , and everyweek mobility anchor configuration is lost on some WLC having code 7.4.100.0.

I am having this exact same problem. I am running 7.3 on 5508 WLC. My remote site LAP's are using Flex (HREAP). The initial access point that my laptop associates to connects with no problem, as soon as I wander out of range of the initial LAP and into the area of another access point, I lose data connectivity. The was validated like the original post as I start a constant ping on the LAN and watch as the ping latency increases and then ping replies stop. The only way to correct the problem is resetting of the wireless adapter on the laptop. Side note my DroidX has no problem wandering from AP to AP.
Laptop: Windows 7 32bit
I then returned to my home site and test where I have a secondary controller and the LAP's are configured for local mode, no problems roaming from access point to access point. Validated with constant ping test. The pings drop for a second and re-
continues as the laptop reconnects.
**Edit: I am going to try the removing the DHCP Addr. Assignment required option, and report that back to the TAC engineer.
Message was edited by: Michael Dunki-Jacobs
**Edit Solved:***
The problem is in deed solved by turning the "DHCP Address Required" but why?

Changing getImage MTU (receive window)

Hi,
I have an application (snippet below) that receives image, runs fine, but like a dog when the network latency increases.
The issue is the MTU size, as when it is increased (windows reg) the speed dramatically increases. I can see solutions to this issue if I want to send the image over my own sockets, writing a server application as well.
Can I change the MTU/receive window that getImage uses?
URL MyURL = new URL(imgName);
pic=getImage(MyURL );
Tracker.addImage( pic,imgID++);
Tracker.waitForAll();
Thanks
Maddiman

Can I change the MTU/receive window that
getImage uses?No, because getImage() doesn't use an MTU/receive window. It just does I/O via the URL. You can't change the MTU or receive window of a URL either. You can do it for a Socket via Socket.setReceiveBufferSize(), or there are Windows registry tweaks to do it globally. If you're on Windows you should definitely do one or the other because the default of 8k is pathetically small.

Latency increase

Similar Messages

Maybe you are looking for