“Details: The server took too long to respond. (The operation has timed out)”

[LSalt]

I installed ESXi 4 on a development PC\Server successfully. It has a Intel PRO/1000 GT Adapter. I can ping it and I can successfully ping from it. When I try to launch the VSphere client for the first time (from 3 different PC’s) I get the following message:

“Details: The server took too long to respond. (The operation has timed out)”

When I check the ESXi host logs I have the following errors in conjunction with my attempt to access the new host.

Ticket issued for CIMON version 1.0 user root

Ticket issued for CIMON version 1.0 user root

SSLStreamImpl BIORead timed out

SSLStreamImpl DoServerHandshake SSL_accept failed with BIO Error

SSL Handshake timedout for stream 192.167.1.150, blacklisting it for 3000 ms

I have tried the following, so far:

1) Uninstalled and reinstalled both host and clients

2) Made changes to DNS on the DNS server

3) Changed network config on the VMware Host (static, dynamic, host name, etc.)

4) Messed with NIC settings on the PC Clients

5) Google the heck out of the errors

6) Tried CLI Client

7) Tried HTTPS://Host IP address

😎 Tried using the

9) Tried using host name

10) Tried 3 different clients

11) Tried to add to existing VCenter Server environment

12) Tried from USB Boot

13) Moved the /etc/vmware/ssl files and restarted the server

14) Followed KB 1003870 Diagnosing the Virtual Infrastructure Client when it fails to connect to an ESX host

15) Turned off Firewall on Client PC

16) Telnet x.x.x.x 443 seems to work

If you have any suggestions I would appreciate them.

[Dave_Mishchenko]

I would enable SSH (see www.vm-help.com) then run tail -f /var/log/messages while you try to connect and post the output back here.

Dave

VMware Communities User Moderator

New book in town - vSphere Quick Start Guide -http://www.yellow-bricks.com/2009/08/12/new-book-in-town-vsphere-quick-start-guide/.

Do you have a system or PCI card working with VMDirectPath? Submit your specs to the Unofficial VMDirectPath HCL - http://www.vm-help.com/forum/viewforum.php?f=21.

[LSalt]

Below are the SSH results you requested and a few additional tests I tried.

1st Test:

Sep 4 20:18:34 Hostd: 2009-09-04 20:18:34.549 4EEB1B90 verbose ‘Cimsvc' Ticket issued for CIMON version 1.0, root

Sep 4 20:18:58 dropbear12504: Child connection from 192.167.1.150:53979

2nd time after reboot:

Sep 4 20:42:19 Hostd: 2009-09-04 20:42:19.253 5BF4AB90 verbose ‘Cimsvc' Ticket issued for CIMON version 1.0, root

Sep 4 20:42:21 dropbear9521: Child connection from 192.167.1.150:54015

Sep 4 20:43:49 Hostd: 2009-09-04 20:43:49.399 5BEC8B90 verbose ‘Cimsvc' Ticket issued for CIMON version 1.0, root

Sep 4 20:44:46 Hostd: http://......................................info ‘ha-license-manager' Update: 85893 minutes left in eval period

Sep 4 20:44:46 Hostd: http://......................................verbose ‘DvsManager' PersistAllDvs Info Called

Sep 4 20:44:58 dropbear9521:exit before auth: error reading: connection timed o

I also tried Telnet "23"-

I got the below message on my windows screen:

Vmware VMvisor <Styx> on vm.aits.com

VMkernel 4.0.0, #1 SMP Release build-171294 Kim 11 2009 12:44:08

Password:

VMware Log:

-Session opened

-Logon root "unknown"

I also tried Telnet on 22, 443 & 80:

Telnet 22 = Windows SSH-2.0 -dropbear -0.51 and in ESXi log I get dropbear

Telnet 443 = Windows blank and ESXi Log get the same messages as I do trying to connect through the VM Client and VM CLI listed in the original question abo ve, "SSL error/blacklisted client IP"

Telnet 80 = Windows blank and ESXi log I get HTTP transaction failed, closing connection

[LSalt]

I continue to diagnosis the problem....

I was concerned that maybe ESXi was seeing the onboard NIC which was disabled in the BIOS. Therefore, I excuted esxcfg -nics -l and it showed only my Intel card. I change the setting to 10MB\Half Duplex in hopes this could fix the problem. It did not.

Next I checked the Message Log and I found a few curious references!

HostSvc:: AutoStartManager retrieve autostart order

Object name"FirewallSystem" failed

I thought I read in a seperate Forum article that ESXi did not have an internal Firewall. Could this be my problem or is this just a invalid comment in the log? How would I check the settings through the console? However, since the I changed the speed settings I am getting something I don't recall seeing before. When I try to access the ESXi host from a URL I get the standard page "Problem with WebSite security Certificate". Further, the Windows IE8 TAB on this page shows "Certificate Error: Navagation Blocked". I added the site to my trused sites, turned off the "Warn about certificate address mismatch" in the IE8 Advance settings. Still didn't work. Below is what I see in the log. The error is smiliar to the original "timed out error", but there is now more details.

Here are the ESXi log entries:

SSLStreamIMPL::DOSERVERHANDSHAKE(D83A9A8) SSL_ACCEPTED FAILED WITH UNEXPECTED EOF

HOSTD:[http://..........WARNING {PROXYSVC|http://..........WARNING {PROXYSVC] SSL HANDSHAKE FAILED FOR STREAM TCP( LOCAL = 192.167.1.150:60159, PEER = 192.167.1.15.:443), ERROR = SSL EXCEPTION: UNEXPECTED EOF

HOSTD: SSL STREAMIMPL:: BIOREAD (d83a9a8) timed out

This sure looks like a ESXi Firewall issue (off on the PC client). And why would ESXi try to return to windows using a random port "60159" instead of 443?

Not sure what more I can try and will wait for any suggestions before proceeding again. Thx

[JBeharry · Accepted solution]

How much memory is in your system?

I am getting the same error but changes the RAM resolves.

MB supports 4 modules up to 8GB each.

With 2 x 1 GB installed 4 ESXi was stable

With 2 x2 GB - 4 GB it errored as you stated

With 1x2GB plus 2x1GB - still timing out

With a 1x2 plus 1x1 (total 3GB) - the system is stable again.

Note that under all circumstances the VM Host console showed a valid IP.

But currently over 3GB and it is unreachable from non-console.

Not sure if this will relate to your issue.

Maybe it will help.

[LSalt]

That solved my problem.......!http://communities.vmware.com/images/emoticons/laugh.gif!

All I can say is WOW and very odd!!!

I originally had 4x1=4GB on my server when I could not connect with the VClient. I reduced, as you suggested to 3x1=3GB and the Windows VClient connected without even a slight hesitation.

I am glad this is only a test box. 3 GB total RAM would be very limiting in a production environment.

Thank you!!!

[JBeharry]

Well I would actually call reducing ram a workaround not a fix...

If I find a way to get above 3GB on mine - I will let you know. Would you kindly do the same for me?

Thanks

John Beharry

[lkx2]

Hi guys.

I have had exactly the same problem with M/B Asus M2NPV-VM and resolved it by disabling "memory hole remap" or something like this in the system BIOS. So the system has 3.2Gb of visible memory out of 4.

I'll try to do some experiments but not sure if they will be helpful. BTW Asus P5Q-Pro works like a charm with 8 Gb of RAM. Maybe it depends on chipset.

[ElCuente]

Hello

I'm getting the same problem in a ESXi server with 32G of RAM, but i can't disable the memory, has anyone advanced in this issue?

[Kooby]

Here's my story.

VMware Server 2 which was going 'alright' decided last night that the Web Interface wouldn't work on a certain computer in IE. The Login boxes were replaced with scroll bars and the Web Interface was just unreliable elsewhere so I thought I would have a go with ESXi. After swapping cpus and motherboards around and purchasing an Intel NIC it seemed that everything was going well. I then noticed that it couldn't get an address via DHCP and when I configured a static IP Address I could ping it but that was the limit of that. I then reduced the ram from 6 GB to 2 GB and SURPRISE it worked...

What a day...

[Basher590]

Hey,

Has anyone found a solution for this problem yet? I am having exactly the same issues. I am using a Dell PE860 which is not supported under ESXi 3.5 and I'd really like to get ESXi 4 running on this box.

Cheers

Chris

[ejward]

I'm having this issue with one host. I have a Dell blade center with 12 M610 blades. Each has 48GB of RAM. One blade is disconnected from VC. I'm running ESXi v4 and have applied all latest patches (Not 4.1 though). I have VMs on this host that are up and running fine so I am limited in what I can do to fix this issue. Any ideas?

[LSalt]

Well I have discovered work around for my Dev Vmware Server that will allow me to see the additional memory. However, I still cannot access the Host directly from the VSphere client. Here is what I did.... I upgraded to 4.1, installed VCenter on a VM, setup the VM to start when the Hosts Starts, shutdown the Host and installed the additional memory. Now I can indirectly access the Host with the additional RAM using the VSphere Client. However, I am accessing the Host by logging in to the VCenter VM.