[Netarchivesuite-devel] [Fwd: [NS-114] Comment: FR1773: treatment of "harvesting aborted" status]

[Colin Rosenthal]

I'm forwarding this to netarchivesuite-devel for general discussion.

In the most recent implementation,
the behaviour is this. If a harvest is stopped either from the heritrix 
GUI, or by killing heritrix, or due to
inactivity timeout or due to a fatal error then all domains in the 
harvest will be marked "harvesting aborted".
On starting a new snapshot harvest based on this harvest, these domains 
will not be included automatically.

Is this behaviour acceptable?

The problem Søren identifies below is that we have no simple way to 
distinguish between harvests stopped
by manual intervention and those stopped due to inactivity timeout.

My own opinion is that at least "harvesting aborted" gives the right 
message in the GUI. The previous
behaviour was that these domains were marked as "domain completed" which 
is downright misleading.

--
Colin


-------- Original Message --------
Subject: 	[NS-114] Comment: FR1773: treatment of "harvesting aborted" 
status
Date: 	Fri, 23 Oct 2009 14:00:33 +0200
From: 	Søren <svc at kb.dk>
To: 	Colin Samuel Rosenthal <csr at statsbiblioteket.dk>
References: 	<r128.d1256280029772 at cruciblethreadindicator>



NS-114 commented(see http://kb-prod-udv-001.kb.dk:8060/cru/NS-114#c1650 ):-

Another issue is, that the text "Heritrix terminated by operator" is also written, when HeritrixLauncher terminates the harvesting due to inactivity.
And therefore not only when the user terminates the harvest from the Heritrix GUI

---

ID: NS-114 http://kb-prod-udv-001.kb.dk:8060/cru/NS-114/review

Title: FR1773: treatment of "harvesting aborted" status

Statement of Objectives:
See https://gforge.statsbiblioteket.dk/tracker/?func=detail&group_id=7&aid=1773&atid=108

State: Review

Author: Colin
Moderator: Colin
Reviewers: (1 active, 0 completed*)
     Søren