Surrogate error in Zope component architecture leading to an OOPS

Bug #52574 reported by Diogo Matsubara
8
Affects Status Importance Assigned to Milestone
BlueBream
New
Undecided
Unassigned
Launchpad itself
Fix Released
Critical
Unassigned
Zope 3
Won't Fix
Undecided
Unassigned
zope.interface
Fix Released
Undecided
Unassigned

Bug Description

<SteveA> it looks like a combination of an edge case in the component architecture (zope adapter registration stuff) and perhaps a leaky sqlobject cache. BugTask has some interesting stuff in its _init where it uses alsoProvides to stick extra interfaces on an object
<SteveA> it's possible that was an edge case of some sqlobject problem combined with a component architecture "being too paranoid" problem

OOPS-190B176, OOPS-193C61, OOPS-199B34, OOPS-309C627, OOPS-380B21, OOPS-384C1271, OOPS-394D863, OOPS-410D744, OOPS-464A218, OOPS-511D246, OOPS-541D1419, OOPS-542A954, OOPS-564E3537, OOPS-570A6278, OOPS-579B744, OOPS-584C490

  Module zope.app.traversing.adapters, line 172, in traversePathElement
    traversable = ITraversable(obj, None)
  Module zope.interface.interface, line 675, in __call__
    adapter = self.__adapt__(obj)
  Module zope.interface.interface, line 742, in __adapt__
    adapter = hook(self, obj)
  Module zope.app.component.hooks, line 96, in adapter_hook
    return siteinfo.adapter_hook(interface, object, name, default)
  Module zope.interface.adapter, line 413, in adapter_hook
    factory = self.lookup1(providedBy(object), interface, name)
  Module zope.interface.adapter, line 401, in lookup1
    return self.lookup((required,), provided, name, default)
  Module zope.interface.adapter, line 333, in lookup
    byname = s.get(provided)
  Module zope.interface.adapter, line 253, in get
    self.clean()
  Module zope.interface.adapter, line 159, in clean
    base.unsubscribe(self)
  Module zope.interface.adapter, line 280, in unsubscribe
    raise KeyError(dependent)
KeyError: <Surrogate(<zope.interface.Provides object at 0x2aaacb143c90>)>

Recently: OOPS-1012G3792

description: updated
description: updated
Revision history for this message
Diogo Matsubara (matsubara) wrote :

Recent OOPS-234D10

description: updated
Revision history for this message
Diogo Matsubara (matsubara) wrote : Re: KeyError exception raised leading to an OOPS

OOPS-246A103

Revision history for this message
Diogo Matsubara (matsubara) wrote :

OOPS-256D194
OOPS-257C434

description: updated
description: updated
description: updated
description: updated
description: updated
description: updated
description: updated
Revision history for this message
Christian Reis (kiko) wrote :

It happens regularly, but not frequently enough (0-2 times a day) -- and very hard to reproduce.

Changed in launchpad:
importance: Undecided → High
description: updated
Revision history for this message
Bernd Dorn (dobee) wrote :

i had also a similar exception in the interface package, but only when using twisted, with zserver the error disapeared

Revision history for this message
Steve Alexander (stevea) wrote :

Launchpad is currently using ZServer and not Twisted, so I guess it's a secondary effect that you didn't see it with zserver but did with twisted.

description: updated
description: updated
description: updated
Revision history for this message
Christian Theune (ctheune) wrote :

Waiting for a reproducable test case.

Changed in zope3:
status: New → Incomplete
Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for Zope 3 because there has been no activity for 60 days.]

Steve Alexander (stevea)
Changed in zope3:
status: Invalid → Confirmed
Revision history for this message
Diogo Matsubara (matsubara) wrote :

Lowering Launchpad task importance, since this is irreproducible and doesn't happen that frequently

Changed in launchpad:
importance: High → Medium
Revision history for this message
Curtis Hovey (sinzui) wrote :

The switch to storm may have increased the chance for this error to occur. I see seven occurrences in the past two days on lpnet and edge:
OOPS-916EA14, OOPS-916EA15, OOPS-916C1893, OOPS-916C918, OOPS-917C1270, OOPS-917EA12, OOPS-917EA13

Revision history for this message
Ursula Junque (ursinha) wrote :

Recently: OOPS-937B2869

Revision history for this message
Ursula Junque (ursinha) wrote :

OOPS-939G1934, OOPS-939G1935

Revision history for this message
Ursula Junque (ursinha) wrote :

OOPS-940E1412

Revision history for this message
Ursula Junque (ursinha) wrote :

OOPS-949B2673

Revision history for this message
Ursula Junque (ursinha) wrote :

OOPS-956E206

Ursula Junque (ursinha)
description: updated
Gary Poster (gary)
Changed in launchpad-foundations:
importance: Medium → Low
status: New → Triaged
Revision history for this message
Tres Seaver (tseaver) wrote :

Can one of the LP guys report on what version of zope.interface this problem occurs? The code path isn't present in the current trunk or the latest releases (3.5.x).

Changed in zope3:
status: Confirmed → Won't Fix
Changed in zope.interface:
status: New → Incomplete
Revision history for this message
Robert Collins (lifeless) wrote :

We're currently dealing with a backlog of oopses making this low frequency one rather hard to look at; once we catch up we'll certainly try to see if this still happens and if it doesn't close the bug.

Changed in launchpad:
importance: Low → Critical
Revision history for this message
Tim Penhey (thumper) wrote :

I'm marking this as fixed released. Since the original report, we have upgraded to zope.interface 3.5.2 and we haven't seen this error at all in 2011. I stopped looking after that.

Changed in launchpad:
status: Triaged → Fix Released
Tres Seaver (tseaver)
Changed in zope.interface:
status: Incomplete → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.