Jump to content
Welcome to our new Citrix community!

LB Vserver objects stuck in "UNKNOWN" state?


Josh Hester

Recommended Posts

Hi, folks!  Long time no talk! 

 

I just upgraded my VPX instances from 12.1 55.18 to 13.0 71.44.  These instances live on older SDX 11520 appliances that are still running SVM version 12.1 51.19 (yea, I know).

 

Everything with the upgrade itself went fine using ADM, and work on the instances since then  -- building new services, changing existing ones, etc. -- has had no issue... until this week.  We noticed that when I try to add any new LB virtual server objects to one of the instances, their starting state is "UNKNOWN," and nothing I do either via the CLI or admin UI changes it.  This happens for non-addressable virtual servers (this is what we noticed first, as we were adding new services to a content switch), addressable virtual servers, and virtual servers of various protocols/types (tested so far: HTTP, SSL_BRIDGE, RDP, just for a wide sample).

 

A few things I've noticed since discovering this:

- Type, addressability of virtual server doesn't matter, as noted above

- I cannot Disable the affected objects in the UI, but I can disable/enable the affected objects in the CLI

- Toggling the object does not change anything

- Being stuck in an "UNKNOWN" state doesn't seem to affect its availability for content switching or actual traffic flow.  For example, adding the affected object to a content switch tests just fine -- it load balances across the servers, and I can validate it by toggling it down in the CLI, seeing the HTTP 503, toggling it on, and refreshing to see the content. 

- Testing the affected object successfully still shows 0 for all stats on the object

- None of my other instances that were upgraded to this same version are experiencing this issue

 

I'm kinda' stumped on this.  I've been pouring through the release notes for this version for a bit, but I've not found anything that addresses.  I took a support file, and I'm now looking at scheduling an HA fail-over so I can reboot the instance just to see if that "shakes something loose."  Has anyone else encountered this? 

 

Thanks!

Link to comment
Share on other sites

  • 1 month later...

I see the same state UNKNOWN only for the "NO-MONITOR" services https://docs.citrix.com/en-us/citrix-adc/current-release/load-balancing/load-balancing-advanced-settings/no-monitor-service.html .

 

From what you say that it works I am starting to think that this is because of the upgrade and not a bug (some kind of version change). You may test checking and enabling monitoring on the services of the Virtual servers that are in UNKNOW state and  check if "monitoring for LB Vserver on a CS Vserver" is enabled or not and goggle it(see below as this option may have changed because of the upgrade) .

 

https://support.citrix.com/article/CTX208818

Link to comment
Share on other sites

  • 4 months later...
On 4/22/2021 at 4:52 AM, Nikolay Dimitrov1709159278 said:

I see the same state UNKNOWN only for the "NO-MONITOR" services https://docs.citrix.com/en-us/citrix-adc/current-release/load-balancing/load-balancing-advanced-settings/no-monitor-service.html .

 

From what you say that it works I am starting to think that this is because of the upgrade and not a bug (some kind of version change). You may test checking and enabling monitoring on the services of the Virtual servers that are in UNKNOW state and  check if "monitoring for LB Vserver on a CS Vserver" is enabled or not and goggle it(see below as this option may have changed because of the upgrade) .

 

https://support.citrix.com/article/CTX208818

 

Hi, Nikolay!

 

This isn't happening to services or service groups -- this UNKNOWN state shows on the LB virtual servers, even on virtual servers whose bound services are actively health monitoring their back-end servers.  The state of the service or service group seems to have no bearing on the virtual server when it's stuck like this.  It's also not specific to Content Switching virtual servers -- most of the virtual servers affected by this are standard LB virtual servers (both addressable and non-addressable).

 

Another thing I've noticed in the time since this problem has manifested: it does seem to impact GSLB's ability to distribute VIP's.  While actual traffic to the VIP for a stuck virtual server isn't impacted, the GSLB service effective state remains DOWN.  I assume this is because GSLB can't tell what the actual state of the object is and fails closed/DOWN instead of failing open/UP by default.  This meas that these new virtual servers that are stuck in an UNKNOWN state are messing with GSLB name resolution for these services.  How frustrating! >_<

 

I honestly don't know what else to do besides see what support can do.  I can't find any references to known issues or bugs of this issue, and even rebooting the instance only fixes objects that already exist at the time of the reboot.  I'll update this post once I've had a chance to get Citrix Support to look at it with me.

Link to comment
Share on other sites

  • 1 month later...

Hi, folks!

I just wanted to put a pin in this.  I ended up just upgrading all my VPX's to the latest stable release of 13.0 (13.0 82.45). That upgrade went through last night, and as of now, the issue of new LB virtual servers showing as "UNKNOWN" appears to be resolved.  I've made a few of them in the time since the upgrade, and all of them accurately show either DOWN or UP depending on the configuration.

 

So ... I guess if anyone else sees this happen, upgrade your firmware?  ?‍♂️  It's not a satisfying answer, honestly, but that's how it goes sometimes. 

 

Thanks!

  • Like 1
Link to comment
Share on other sites

  • 3 months later...
On 8/25/2021 at 12:57 PM, Josh Hester said:

 

Hi, Nikolay!

 

This isn't happening to services or service groups -- this UNKNOWN state shows on the LB virtual servers, even on virtual servers whose bound services are actively health monitoring their back-end servers.  The state of the service or service group seems to have no bearing on the virtual server when it's stuck like this.  It's also not specific to Content Switching virtual servers -- most of the virtual servers affected by this are standard LB virtual servers (both addressable and non-addressable).

 

Another thing I've noticed in the time since this problem has manifested: it does seem to impact GSLB's ability to distribute VIP's.  While actual traffic to the VIP for a stuck virtual server isn't impacted, the GSLB service effective state remains DOWN.  I assume this is because GSLB can't tell what the actual state of the object is and fails closed/DOWN instead of failing open/UP by default.  This meas that these new virtual servers that are stuck in an UNKNOWN state are messing with GSLB name resolution for these services.  How frustrating! >_<

 

I honestly don't know what else to do besides see what support can do.  I can't find any references to known issues or bugs of this issue, and even rebooting the instance only fixes objects that already exist at the time of the reboot.  I'll update this post once I've had a chance to get Citrix Support to look at it with me.

 

 

Did you end up finding a resolution?  Just came upon the same situation.

 

Thanks.

 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...