In ESXi 5 very slow network performance

Similar Questions

• Very slow network performance

  I have a Dell T610 running ESXi 4 with a pair of built in maps network gigabit Broadcom BCM 5709, to which I've also added a gigabit dual port Intel 82576 NIC they are all plug a Cisco 3560 24 ports switch. There are a couple of virtual machines on this computer and there was near zero, a network, or activity of the CPU during the following tests as they were made during off-peak hours.

  I noticed a few updates takes a long time to download and, after completed and restarting the virtual machines, I decided to perform a speed test that indicates the very slow network performance. But the performance is slower on one NETWORK card to another. Reduce the ports available to virtual machines on a single NETWORK card, that's what I got.

  Using Speedtest.net as a gauge running on a dedicated XP Guest:

  The Broadcom and Intel befriended 5Mbps showed and sometimes 10 Mbps.

  Only Broadcom showed between 13-18Mbps from one trial to another.

  Intel showed only 4-5 Mbps. never more than 5 Mbps!

  Physical work stations connected to the same switch showed 45Mbps repeatedly!

  All ports were originally configured as auto negotiate. I tried locking time ESXi and the Cisco switch to 1 GB Full duplex with no significant difference in the overall performance.

  Has anyone seen this? Have you found a solution?

  I got a similar promblem - found that the VM - NIC has been set at the 'flexible' type and VM-Ware tools were not properly installed.

  As I was not able to install (linux, no gcc), I've removed the NETWORK card and put in again as an adapter 'e1000 '. From here, I had transferrates as expected ;-)

  I hope this helps...

• Workstation 9 - slow network performance between the host and guest

  I am running windows 8 on the host and installed by 9 trial. Then installed windows 8.1 as a comment system.

  I'm a slow, very slow network between the two.

  If I send files to and from the computer with another guest, transfers are normal, but if I try the same files between the host and the guest, they are really slow.

  I use a bridge connection.

  I read elseware to disable the following on the guest card:

  Large Send Offload v2 (IPv4)

  Large Send Offload v2 (IPv6)

  But that did not help.

  Thoughts?

  Maybe then try Hyper-v, it help you better.

  report this incident to Vmware if you can

  Concerning

• Slow network performance for non - vm with ESXi 4.0 machines

  I have ESXi 4.0 running on an IBM x 3850 M2 with 2 6-core Xeon, 64 GB RAM, 1.2 to SAS RAID 0 and 2 Gbit NIC connected to the 100 MB switch.  Two physical network cards is configured to use a vSwitch.  The management network use vmnic0 with vmnic1 as before.  The virtual computer network use vmnic1 with vmnic0 as before.

  I created a half-dozen VMs with VMware Tools installed, a certain Win XP, Win2003 Server.  Access to the network between the virtual machines is ok (_{270Mbits/s), but access to external guests is very slow (}2.5Mbits / s).  Here's the output of iperf, first between two virtual machines (tnealxpvm and rptkerbtest1), and then between a virtual machine and an external host (tnealxp03).  I guess the network access VM-on-VM is probably in memory and touch even the network hardware?

  C:\tneal & gt; iperf.exe w 128 k rptkerbtest1 - c

  -

  ---

  Client that connects to rptkerbtest1, port TCP 5001

  TCP window size: 128 KB

  -

  ---

  9.42.124.124 local port 1159 connected with 9.42.124.75 port 5001

  Bandwidth transfer interval

  0, 0 - dry 10.0 321 MB 269 Mbit/s

  C:\tneal & gt; iperf.exe w 128 k tnealxp03 - c

  -

  ---

  Client that connects to tnealxp03, port TCP 5001

  TCP window size: 128 KB

  -

  ---

  9.42.124.124 local port connected 1336 with 9.37.207.29 port 5001

  Bandwidth transfer interval

  0, 0-11, 4 s 3.35 MB 2.47 Mbps

  In the other direction (external host to VM) network speed is even worse.

  D:\tneal\downloads & gt; iperf.exe w 128 k rptkerbtest1 - c

  -

  ---

  Client that connects to rptkerbtest1, port TCP 5001

  TCP window size: 128 KB

  -

  ---

  9.37.207.29 local port 18053 connected with 9.42.124.75 port 5001

  Bandwidth transfer interval

  0,0 to 13.7 sec 984 kb 587 Kbits/sec

  I also run iperf on the hypervisor to try to determine if the problem is with the installation of the physical network.   Here, the speed is good/expected with a 100Mb connection.

  / tmp/Eddy #. / iperf w 128 k - c 9.37.207.29

  -

  ---

  Client that connects to 9.37.207.29, port TCP 5001

  TCP window size: 129 KB (WARNING: requested 128 KByte)

  -

  ---

  9.42.126.252 local port 53063 connected with 9.37.207.29 port 5001

  Bandwidth transfer interval

  0, 0 - dry 10.0 112 MB 94.2 Mbps

  Suggestions for changes to configuration or other tests would be greatly appreciated!

  -tom

  Hello

  No reason to hardcode them guests at 100 Mbps.  Configure your guests to autodetect speed connection and duplex.  It's ok if your report speed/duplex of link guests that are different from your physical network.

  Try to configure your NIC Host (ESX) autodetect. Check that they autodetect to 100/Full.  Some older switches have problems with autodetecting link speed/duplex, but I prefer to configure NIC gear to detect automatically if possible.

• vSphere Home Lab: very slow network speed, I'm desperate

  Hello

  After many trials that I have finally found the right hardware build my vsphere home laboratory.

  This is my setup:

  ESXi1

  ASUS P6T Deluxe V2

  I7 920

  24 GB OF DDR3 MEMORY

  2 x Intel Pro 1000 GT Desktop Ethernet adapter

  2 ESXi

  i5-2500

  16 GB OF DDR3 MEMORY

  1 x Intel Pro 1000 GT EthernetDesktop adapter

  NAS

  ASUS P7H55-M

  Intel Core i3-540

  Windows Server 2008 R2 with Starwind iSCSI

  Realtek Gigabit Ethernet

  I have a Netgear GS108 where I plugged all the ethernet cables over my ADSL Modem, a Linksys WAG320N

  Everything is OK apart I have a speed very very slow in the network; for example ISO 2k8r2 server transfer I take 30 minutes

  After many trials, I decided to start from scratch, I just powered on my ESXi2 machine and reinstalled ESXi 4.1U1 from scratch and I installed the Client vSphere for my Client PC which have a Gigabit Ethernet card and is also attached to the switch Gigabit however, speed is still the same, 2/3 MB / s.

  The strange thing is that also the same speed very slow, I created a Server 2008 VM and installed iperf to landmark of network and I get about 600 / 700 Mbps, so I really don't know what I can do.

  I replaced the ethernet cables, also tried to connect the cables directly to the modem and also I tried with another switch, but the result is always the same.

  He is now a cable or switch problem because I tried to create an iSCSI drive and transferred a file of 10 GB of my Client PC to the iSCSI disk and I get about 120 MB/sec so is full Gigabit network speed.

  Also, I installed FileZilla FTP Server on the ISCSI server and tried to download and upload a file of 10 GB and the speed is never 115 / 120 MB/s.

  Really, I don't know what I do that, I'm really desperate.

  Thank you very much for your help.

  Routers are NOT created equal.  There is a business reason switches cost $10,000 - $20,0000 dollars.

  Netgear router / switch is just a very BASIC.  It's not classy company.  That's your problem.  Get a better switch that you don't need to spend $10,000 but probably a few hundred... a GOOD show for home switch...

• Very slow network with 10 Workstation and Windows 8 / 8.1 Windows

  Hello

  I come to you because I have a problem with the network with 10 Workstation and Windows 8 / 8.1

  In fact, if I copy files from or to a virtual machine with my host of Windows 8 by sharing Windows (SMB / CIFS), it is very very slow and unusable.

  Example: My host is a Windows 8 x 64 and Windows 8 x 64 host: transfer speed is about 100 or 200 KB/s and often 0 KB/s (transfer seems to be paused).

  With a guest Windows 8.1, same problem.

  After many tests, I also tried with Windows PE 5 (same kernel as Windows 8.1): same problem, unusable.

  But, miracle, if I use "Utility MSI - X OFF" Intel, I reboot and the problem disappears.

  There:https://downloadcenter.intel.com/Detail_Desc.aspx?lang=eng & DwnldID = 18946

  Someone in this community, said that it have the same problem and a workaround solution was not to use Intel e1000e but e1000 then: I must admit, it works.

  With e1000, speed is around 30 to 60 MB/s and stable.

  But, by default, 10 Workstation e1000e value when I create a new VM Windows 8.1, so it's a problem... Do not manually change for each new virtual machine and explain this to my students.

  Please, you can reproduce this bug and fix it?

  Thank you very much.

  New workstation 10.0.2 fixed this bug. Perfect.

• Slow network performance

  Hello. Since version 6.4.0 I have very slow write speed. Now I launch last 6.4.2 RC1 and I have the same problems. My antivirus is disabled because I bought my readynas. While I can copy ReadyNAS with speed 20-30mbps under gigabit lan when I try to write something via lan I get 5-7mbps speed. Is there a solution? I must say that before 6.4.0 version I've had close read and 70mbps write speeds!

  Mastoul wrote:

  No I don't because it worked before, so I didn't touch it.

  Try without regrouping and see if the behavior changes.

• ThinApped Firefox 11 slow network performance

  Hi all

  We have thinapped firefox 11 with thinapp version 4.7. However, we have serious network performance problems when using firefox. In the thinapp firefox on speedtest.net, we get 1000ms ping ~10mb/s download and 3 MB/s in upload.

  On the native IE9, we get 100 ms ping ~95mb/s 39 MB/s in upload and download.

  At first I thought it was a sandbox problem, as we have our sandbox in %APPDATA%\microsoft\windows\sendto that is redirected to a network drive. I changed this option to use the current directory of the thinapp, but this made no difference. I also tried to add the "DisableRegistryTransaction = 1" option, but it also made no difference.

  The ThinApp was captured on a clean XP VM and I test it on our version of W7. I tried this on our XP build legacy and receive the same results.

  Does anyone else have this problem? And are there any suggestions? For the moment, we are looking to put firefox in our basic build or delivery through SCCM.

  Val

  See here for ThinApp Antivirus recommendations:

  http://blogs.VMware.com/ThinApp/2012/02/antivirus-scanning-in-a-VMware-view-virtual-desktop-environment-that-includes-ThinApp-virtualized-applications.html

• ESXi Server very slow

  I am managing an ESXi v3.5 and it is slower than molasses!

  All VM is running very well, as there is no problem, but connection and try to manage the host ESXi expires more often it connects.  Under performance - & gt; system is an alert under "criticism for unspecified Int 1 - Bus say unrecoverable error."  And someone in front of me seems to have been managing this host with a Virtual Center Server and he thinks he always handled.

  So, a few questions:

  (1) what is the error system supposed to tell me?

  (2) a reboot will cut the ghost Virtual Center Server?

  (3) any ideas on slow?  A reboot that will correct perhaps?

  b

  Good to know. Look, this isn't official or anything, but I have seen sfcbd significantly reducing the performance of the E.C. on ESXi. You need to upgrade to the latest version of ESX (U4) so I think that you will get rid of this.

  Services can be kept low without problem. The only thing you lose is a kind of material maintenance for the 3rd party software.

  If you can, check this discussion as resolved so I can try to win a prettier icon

• x 220 slow network performance

  I have a x 220 windows 7 32 bit and with cable and wireless I get extremely slow performance. The chip is an Intel 81579LM and copy files over the network via the cable network for example I get only 800 Kbps. If I plug my desktop PC into the same switch I get something much faster. So the problem is definitely on the 220 x. I tried various driver updates, but nothing seems to fix. I ran perfmon and CPU and disk read/write queues are not a problem. Where should I go from here?

  Roger

  roger456 wrote:

  If I plug my desktop PC into the same switch I get something much faster

  .

  You use the same Ethernet cable between the device and the same port on the switch for the two tests?

  If this is not the case, try to do the test with the help of two Ethernet cables between the devices and the same port of the switch... This would pretty quickly sort if you have an Ethernet wiring related issue - it's pretty common root cause of the problem you describe.

  See you soon,.

  Bill

• Photoshop very slow keyboard performance move + boost

  I have a small (24 MB) file, which has many layers. All of a sudden, to move a single item (vector shape layer line, for example) takes 3-4 seconds to around 10px from the time of the press on keyboard. It's very laggy and agonizingly slow. New PS documents seem to be good. My history and cache is attached to the large and thin.

  
  

  This will be exactly like a very old problem Photoshop had return to the CS6. My latest version of PS 2014 seemed fine.

  
  

  I have an end of 2013 retina MacBook Pro, 16 GB ram, SSD drive with 80% free space.

  
  

  Any ideas?
  

  I think only now would be that you have a LOT of layer comps in the document.

  But I think I've also fixed a few layer of movement performance issues in Photoshop CC and CC 2014.

• Lightroom 3.3 very slow export performance - help?

  Hello

  Last night, I installed the patch 3.3 and I began to edit the value of a day of pictures. Everything seemed fine until I went to export photos. I had only 140 photos, exported as jpg, 800 x 600, 72 dpi, while they were quite small. Normally it would take upt 5 minutes to do, but last night after about an hour, it was only 36% of fact. On top of that my whole system slowed and frozen, and I had to reboot. Export the remaining photos took another hour, so restarting does not solve the problem.

  If there isn't a fix published soon, I have to downgrade to 3.2, but I don't know how to do this. My business is suffering because of this.

  I'm operating a fast processor on a PC with 4 GB of RAM so I should be fine. The performance was excellent until the recent upgrade, and I have not installed anything recently.

  Help!

  Hello world

  I just stumbled upon this very exact problem. Export of lot of 378 points jpg resized and LR 3.3 minutes filled my SSD with 40 GB of data scratch/temp and was only about 150 items in the export process.

  Since I rarely use the 'Limit File Size To' option in the dialog box export jpg, I could quickly identify it as the culprit (at least on my system).

  Turning off this option returns LR 3.3 to its normal operating speed without creating a huge temp files.*

  Hope this indication is useful in some way and report if this workaround works on your system.

• OfficeJet 8720, scan to folder very slow network. any tips?

  I just upgraded to Officejet pro 8600 to 8720.

  The 8720 prints well, but it is incredibly slow for scanning.   Can't believe how slow it is compared to other HP products I've used.

  Is there advice, suggestions on how to speed up scanning?  Or do I need to buy a more expensive model for a faster sweep.

  Please notify.  Thank you.

  Hi @atomarchio,

  I was check in this issue and found a number of things that can help.

  Replace with the IP address of the computer that hosts the folder targets of network for the name of this computer.  For example, if the network path is listed as \\david\8670scans, change it to read \\192.168.1.102\8670scans.

  Turn on the printer and try again.

  Check out this post from @djjaeger82, Officejet Pro 8720 - slow Scan to network folder.

  Let me know how make you out. Thank you!

• HP Officejet Pro 8600 printing very very slow - network printer

  Print very slowly.  For a document of several pages that can be printed a correct page followed by letters and random symbols on the following pages printed on a line at the top of the sheet of paper.

  Printing test page OK.  Using the Windows 7 on an Apple Mac operating system.   Impression was carry out well so far.  Tried stop, check the cartridges etc.  Nothing seems to work.

  I am using Bootcamp.  However this communication of computer printer works perfectly for months until this time.  I have tried everything, checked everything.  Ensured that I cancelled all printing of all laptops (our printer is connected via WiFi at home (Home Office).  Finally, I pulled out the mechanism of double-sided printing see if paper jams exist.  No jam.  Moved to the printer across the room, small room of 6 feet.  Plugged into an outlet different.   Restarted my computer under Mac OS, operating system, and print.  Print perfectly.  Restart Windows via Bootcamp and he continued to print perfectly.   Lost in the morning to try to make it work! But happy now. Can't really say that one thing has got to work.  (perhaps restarting the computer?)

• HP ProLiant ML310e with ESXI 6.0 slow disk performance

  Just a question if this issue was already addressed with a new driver?

  Performance is really bad HARD disk access. Less then USB2.

  Thank you

  Rui

  Nope. Downgrade from version 088 is still necessary.