Wildcard SSL Certificate Request
Status of VIPs
VLM HA Issue
Hi Guys,
I've encountered a few similar issues everytime I setup a VLM HA cluster.
Below is the description:
- Upon setup the 2nd node of the VLM, after retrieving configuration from the HA partner, I perform a reboot.
- After the reboot, the 2nd VLM will attempt to become the MASTER while the 1st node remain as MASTER.
- When both VLM becomes MASTER, they will not be able to ping each other even though they r on the same network.
- Need to perform numerous reboot or go to the extend to reset to default for the 2nd node and reconfig the HA
- Both VLM are configured with static mac address
- On one occasion, my HA is setup properly and working fine, then I attempt to shutdown one of the VLM. The standby VLM will turn MASTER (Working normally). But when I power on the previous shutdown VLM, it attempt to turn itself into MASTER (Now 2 nodes becomes MASTER)
I've tested before for the HA functionalities so I'm puzzled why this issue still occurs.
*Encounter on both HyperV environment and VMware environment.
Would like to know what can be the possible caused? Will there be additional configuration to be perform on VMware Vswitches or physical network?
I understand that in HyperV environment you will require to turn on MAC spoofing but there is no similar settings in VMware except enabling Promiscuous mode
Thanks
HA Issue
Hi Support,
I've encounter on few occasion that my VLM nodes will attempt to turn themselves to MASTER in HA mode.
Both of them are connected to the same network and able to setup properly with shared IP.
But at times they will turn themselves to MASTER and ignore each other presence.
I've configured both VLM with static mac but is there any other settings that needs to be enable on the vswitch? e.g. Promiscuous mode etc?
Understand that the HyperV VLM have mac spoofing but VMware doesn't have it.
Please help
What does "kernel: L7: Decode time failed " mean?
FTP how to?
I'm trying to install FTP services with VLM vers 7.0.8.
here is a description :
2 real servers :
ftp1.domain.com (IP1: 192.168.1.X ) and ftp2.domain.com( IP2:192.168.1.Y)
Both are running on debian7 with Proftpd server ( listen port 21 and passiv mode enabled)
FTP virtual service created as follows :
Service NAME :FTP
Alternate Address :non
Service Type: Generic
Activate Service :yes
Force L7:yes
Transparency :yes
No extra ports
Server Initiating protocols : Normal Protocols
Persitence options :nono
Sceduling Method : Fixed weighting (1000 for ftp1 and 500 for ftp2)
idle connection :none
Use address for server NAT : no
Quality of service : Normal service
Real servers : ftp1 and ftp2
VIP: 192.168.2.X
The status is green and up.
FTP is working when connected directly to the real servers. But not working on VIP .
Do I miss something ?
I really appreciate your help,
Thks guys,
SSL connection taking longer time
users try to connect, we try every aspect on KEMP configuration to solve
this,
and then we found out that if we used HTTP Host Header as persistence
option, the web is doing fine, there is no captcha again to connect to
this website
Then we put the load-balancer into live-production environment (assuming everything is working good)
But, there goes another problem, the users feels it takes time to connect to their website now,
and when we try to put the KEMP aside (users connect to the server
without any load balancer between them), everything goes back to normal
Load balancing application pools IIS 7.5
Hi,
I have 3 instances on 2 real servers, LIVE, TEST and TRAIN. They all use .NET 4 and are in different app pools respectively named.
I have managed to setup load balancing on LIVE app pool by having the health check page set to check mydomain.com/LIVE/login.aspx so that if an app pool on LIVE were to fail, then it would get a .NET error and direct all traffic to the server that doesn't have the failed app pool.
Now my problem is I now need to do this for TEST and TRAIN, because the health check in the VS only checks the live one, then even if I stop the TEST app pool, it will still give some of the users a .NET error and not direct traffic to the working server. I thought I had the answer with sub virtual service so I could create multiple virtual services that connect to the same 2 real servers. This let me setup 3 health checks. But I am having trouble figuring out how to direct the TEST and TRAIN traffic to those sub virtual services (or more specifically use those values for health checking)
I have had a look at the content rules and had a go but I must not be understanding this part of the load master config very well. Any help is appreciated.
Carl
HA Configuration
warning log e-mail
I have received 2 warning e-mail from my kemp load balancer:
Oct 19 07:29:04 KEMP1 kernel: net_ratelimit: 10 callbacks suppressed
Oct 21 14:00:01 KEMP1 checkproc: checkproc: xread error: No such process
What is the meaning of these messages?
VLM1000 Max Ethernet Interfaces
ESP and OWA, session timeouts
Exchange 2010 loadbalanced with two LM2600, version 7.0-8a.
When user logs on through ESP they can work in OWA but, after 5-10 minutes it seems to timeout.
It doesnt logoff like manually it shows the following box when opening a mail for an example:
"Your network connection is unavailable . Please reconnect and try again."
Some users need to restart their computers to be able to reconnect.
Public and Private Computer settings on the Exchange servers are set to 60 min.
Does Load Master supports connection mirroring in HA configuration
CAS Proxying
Hi,
We have 4 Client Access Servers, split 2 per site (let's say site A and site B). 2 of these CA servers are on the same site as the LM (site A), which has a virtual service configured for HTTPS with SSL Offloading and ESP (for all services OWA, ECP, Autodiscovery etc).
For clients who access OWA, who's mailbox is on the remote site (site B.. i.e. the site that doesn't have the CA servers configured as real servers in the LM), will CAS proxying still work from the CA servers in site A to site B?
Also - will I need to configure the virtual directories on the client access servers at site B with no SSL required (as per site A)?
Thanks
Tony
Exchange 2010, Kemp and OS X
Default Gateway
Hi,
is it possible to have one different "default Gateway" for each eth on LM, or do i have to set additional routes?
Kind Regards
mwerner
SSL Certificates Randomly Stop working
Good Morning, for the last year or so we have had
"bizarre" SSL cert issues with the loadmaster. First on the LM 2200 then the virtual LM and
now on our 2600 and we just performed a firmware upgrade to the latest
Vers:7.0-8a and we still have an issue.
Here is what I see. Today, for
example, we cloned an SSL VIP and made a new SSL VIP with the same cert and
when you go to that URL it just hangs. I
replace that valid real cert with the loadmaster Self signed on and then VIP
works again. Now SOMETIMES, you can just
replace the valid one right back and the SSL VIP works, and sometimes it does
not. And randomly some SSL VIPs stop working
and we cannot figure out why. We have
had to RE-IMPORT the cert with a new name at times to get it working. This is very frustrating. Any ideas? Seems to be only with Wildcard certs thought. Anyone else have this issue?
-Wes
Kemp ESP/OWA public vs private
This is currently configured and working quite well.