HCL Traveler PrimeSync threads stuck in native Session.NgetDatabase(), preventing clean shutdown

HCL Domino Domino Forum
1

Domino/Notes Version: 12.0.2FP8
Add-on Product (if appropriate, e.g. Verse / Traveler / Nomad / Domino REST API):
Its Version:
Operating System: Rocku Linux
Client (Notes, Nomad Web, Nomad Mobile, Android/iOS, browser version):

Problem/Query:

Hello everyone,

We have an issue with HCL Traveler.

Periodically, PrimeSync (PS) threads become stuck while opening remote mail NSF databases via NRPC (TCP port 1352).

What we observe:

  • PS threads remain inside lotus.domino.local.Session.NgetDatabase() for more than 200–300 seconds.
  • The affected threads show RUNNABLE, In Native=true, with almost no CPU consumption.
  • The issue occurs with mail databases hosted on different Domino mail servers.
  • Java thread dumps report no Java deadlocks.
  • Traveler sometimes reports Unable to prime sync (Internal Error) while the associated native thread remains blocked.
  • When we try to stop Traveler, some threads do not terminate, causing shutdown timeouts and requiring further recovery.

We also observe HTTP requests such as getStatus and setConfig taking approximately 300 seconds and returning HTTP 409. We have not yet established whether these HTTP delays have the same root cause.

Initial TCP socket snapshots did not reveal significant retransmissions or growing TCP queues, although these snapshots were not sufficient to rule out network issues during an actual hang.

Questions:

  1. Are there known HCL Traveler or Domino issues involving Session.NgetDatabase() hanging during remote NSF access?
  2. Which diagnostic traces would help distinguish an NRPC network wait from an internal Domino API lock or a mail-server-side delay?
  3. Are there specific Traveler or Domino debug settings recommended for investigating this behavior?
  4. Is there a known fix or workaround for native threads that do not terminate during Traveler shutdown?

Any suggestions would be appreciated.

2

Hi ksandrix,

It seems like there are several issues that need to be addressed in your current scenario. You can submit a Support Case to us so we can further analyze it. Thank you.

3

Hi Ksandrix

Does the incident really have a direct connection with the NRPC access to the mailbox, or is that more of a coincidence? I had a similar problem at a customer’s site once. I can no longer remember exactly which step in the Primesync was blocking, but in any case, the SSL deep inspection feature was enabled on the firewall through which the ActiveSync traffic from the Traveler passed, and at some point this interfered with and blocked the PrimeSync. The device was waiting for a confirmation that was being blocked by the firewall. If I remember correctly, the firewall may have been a Barracuda.

Cheers, Daniel