[Condor-users] Flocking problem...

[Rob <spamrefuse@xxxxxxxxx>]

Hi,

I have a problem with flocking between two linux master PCs. Here is what I have in the respective logs:

Linux Master (condor version 7.4.2) with many pool PCs. This is its NegotiatorLog:

02/14 20:22:55 ---------- Started Negotiation Cycle ----------

02/14 20:22:55 Phase 1:  Obtaining ads from collector ...

02/14 20:22:55   Getting all public ads ...

02/14 20:22:55   Sorting 7 ads ...

02/14 20:22:55   Getting startd private ads ...

02/14 20:22:55 Got ads: 7 public and 0 private

02/14 20:22:55 Public ads include 2 submitter, 0 startd

02/14 20:22:55 Phase 2:  Performing accounting ...

02/14 20:22:55 Phase 3:  Sorting submitter ads by priority ...

02/14 20:22:55 Phase 4.1:  Negotiating with schedds ...

02/14 20:22:55   Negotiating with weenee@xxxxxxxxxxxxxxxx at <115.145.158.160:58283>

02/14 20:22:55 0 seconds so far

02/14 20:22:55 condor_read() failed: recv() returned -1, errno = 104 Connection reset by peer, reading 5 bytes from schedd weenee@xxxxxxxxxxxxxxxxx

02/14 20:22:55 IO: Failed to read packet header

02/14 20:22:55     Failed to get reply from schedd

02/14 20:22:55   Error: Ignoring submitter for this cycle

Another Linux Master (condor version 7.7.3) with a user "weenee" as weenee@xxxxxxxxxxxxxxxxx

This one has no pool PCs, but can flock to previous master. This is its SchedLog:

02/14/12 20:22:55 (pid:1649) Using negotiation protocol: NEGOTIATE

02/14/12 20:22:55 (pid:1649) Can't receive negotiation header

Does somebody have a clue what is the problem here?

Is there a condor version problem?

Thank you!

Rob.