January 2002 - Zope - Zope lists

18 Jan '02

Could someone have a look at the following 'Boring' class with the security functionality added (as described in ZopeBook/6.Security and some other products). Could 'security' machinery be broken in Zope-2.4.1 ? It surely doesn't seem to work as adverised, on my machine at least (Debian Linux 2.2, Zope 2.4.1 (source release) python 2.1.0, linux2). Tell me if it works on your installation. Boring.py -------------------------------- __doc__ = "" __version__ = '0.1' import Globals from Globals import HTMLFile # fakes a method from a DTML file from Globals import MessageDialog # provid from Globals import Persistent # makes an object stick in the ZODB import OFS.SimpleItem import Acquisition import AccessControl.Role from AccessControl import ClassSecurityInfo READ_PERM = 'View Stuff' WRITE_PERM = 'Change Stuff' security = ClassSecurityInfo() manage_addBoringForm = HTMLFile('boringAdd', globals()) def manage_addBoring(self, id, title='', REQUEST=None): """Add a Boring to a folder.""" self._setObject(id, Boring(id, title)) if REQUEST is not None: return self.manage_main(self, REQUEST) class Boring( OFS.SimpleItem.Item, # A simple Principia object. Not Folderish. Persistent, # Make us persistent. Yaah! Acquisition.Implicit, # Uh, whatever. AccessControl.Role.RoleManager # Security manager. ): """Boring object. """ meta_type = 'Boring' # what do people think they're adding? manage_options = ( # what management options are there? {'label': 'Edit', 'action': 'manage_main'}, {'label': 'View', 'action': ''}, # defaults to index_html {'label': 'Security', 'action': 'manage_access'}, ) # NOTE: commented out following as it seem to conflict with # 'security.declareP...()' declarations later on # __ac_permissions__=( # what permissions make sense for us? # ('View management screens', ('manage_tabs','manage_main')), # ('Change permissions', ('manage_access',) ), # ('Change Borings' , ('manage_edit',) ), # ('View Borings', ('',) ), # ) def __init__(self, id, title=''): """initialise a new instance of Boring""" self.id = id self.title = title # SECURITY - ==================================================== # here I played with '#'s, then simply tried to access 'index_html' # after each security declaration, # as user 'Anonymous', and noted the results on same line. # 'NOT-WORKING' simply means not working as advertised (allowed access when # it shouldn't, and vice-versa). As you can see, there are too many # 'NOT-WORKING' results. Do you come to similar results? # My conclusion is that security declarations have no effect whatsoever, # whether I declare something, then its oposite, I end up with the same # result. This shouldn't be. security.setPermissionDefault(READ_PERM, ['Stuff Manager','Manager']) security.setDefaultAccess('deny') # <== NOT-WORKING # security.declarePrivate('index_html') # <== NOT-WORKING # security.declarePublic('index_html') # <== OK # security.declareProtected(READ_PERM, 'index_html') # <== NOT-WORKING index_html = HTMLFile('index', globals()) security.declarePublic('manage_main') # <== NOT-WORKING manage_main = HTMLFile('boringEdit', globals()) def manage_edit(self, title, REQUEST=None): " " self.title = title if REQUEST is not None: return MessageDialog( title = 'Edited', message = "Properties for %s changed." % self.id, action = './manage_main', ) Globals.InitializeClass(Boring) --------------------------------

2 1

How do I find which object causes a 'No Authorization header found' error ?
by vio 18 Jan '02

18 Jan '02

Hi, I have a product with many classes, and an access bug. How do I find out which object is causing the bug? The traceback: Traceback (innermost last): File /v/Zope-2.4.1-src/lib/python/ZPublisher/Publish.py, line 223, in publish_module File /v/Zope-2.4.1-src/lib/python/ZPublisher/Publish.py, line 187, in publish File /v/Zope-2.4.1-src/lib/python/ZPublisher/Publish.py, line 162, in publish File /v/Zope-2.4.1-src/lib/python/ZPublisher/BaseRequest.py, line 463, in traverse File /v/Zope-2.4.1-src/lib/python/ZPublisher/HTTPResponse.py, line 581,in unauthorized Unauthorized: (see above) This traceback is not very specific, it says 'Where' there was an error 'inside Zope', but not 'Who' caused it 'inside MY code'. I guess I can go into Zope source and start modifying it a little to make it spit the info I want, but that's not really an acceptable solution, is it (sounds ugly). So again, what I want is to know where the problem is inside MY code, not inside Zope. Any help or tips? Vio

2 1

RE: [Zope] LocalFS and mega-uploads
by larry_prikockis＠NatureServe.org 18 Jan '02

18 Jan '02

...Just to kind of wrap this thing up... Thanks to those who helped me figure this out. As several people have now suggested and confirmed, the problem was the partition I had /var/tmp on filling up during the upload process. Moving /var/tmp to a bigger partition cured the basic problem I was having. However, transferring of extremely huge files still put more of a load on the system than I would've liked. In the end, what I've ended up doing is setting up a separate standalone FTP server, locking it down as much as possible and using an html/script front-end to allow some degree of integration with the rest of my Zope site. It's not quite as nicely integrated as using Zserver's native http/ftp capabilities, but it does the job. > -----Original Message----- > From: Jim Penny [mailto:jpenny@universal-fasteners.com] > Sent: Wednesday, January 16, 2002 7:22 PM > To: larry_prikockis(a)NatureServe.org > Subject: Re: [Zope] LocalFS and mega-uploads > > The /tmp path hypothesis does seem correct. I just started a > megaupload, > and found this--that space went down as the upload > progressed, (as reported > by df -k). But there were no unusual files in /tmp! > > However, looking at lsof's output, I see: > bytes > vvvv > python1.5 32012 zope 24u REG 3,1 56111104 448341 > /tmp/(a)31823.224 (deleted) > > So, it looks like cgi.py is being activated, and that it is stuffing > the upload into a file that it is immediately unlinking. The upshot > of this is that the file is not present to prying eyes. My script > was then reading it from /tmp in 4k chunks. > > This being the case, the best thing to do is to make sure that /tmp > has a LOT of room available. You want to make sure that it is mounted > on a separate partition from the data's destination and preferably > from anything else. Otherwise, your hard drive will have to have > at least twice as much space free (one for the /tmp copy and one for > the destination copy) as the largest expected upload. > > Note, it is fairly conventional to have /tmp mounted on /. Moreover, > many administrators keep / intentionally small. You don't want to > do this here: you can keep / small, but only if /tmp is mounted > elsewhere!

3 2

Unknown entries in PCGI log (Urgent Please help!)
by Steven Grimes 18 Jan '02

18 Jan '02

All, Fri Jan 18 11:48:12 2002 pcgi-wrapper: Failed to start resource (101) failure during start I've gotten the following error in the pcgi.log file. After this error I have to restart the zope process. I've tried searching the mailing list archives with no luck. Anyone have any clues? Also, is there anywhere that I can find a complete list of the pcgi errors? Current configuration: Window NT 4 SP 5 Zope 2.2.0 running through IIS 4 with PCGI. Thanks, Steven Grimes Database Engineer sgrimes(a)acceleratedtechnology.com Accelerated Technology, Inc. 720 Oak Circle Drive East Mobile, AL 36609 TEL: 251-661-5770 FAX: 251-661-5788 www.acceleratedtechnology.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Over Ten Years of Providing Embedded Solutions ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

1 0

Re: [Bug] Formulator 1.0 modifies ZODB on startup
by Dieter Maurer 18 Jan '02

18 Jan '02

e.frerich(a)nord-com.net writes: > >>>Patches (working for Python 2.1 and above) available on > demand. > > May I have the patches? The patch can soon be found via my Zope page <http://www.dieter.handshake.de/pyprojects/zope> Dieter

1 0

Zope as daemon
by Allen May 18 Jan '02

18 Jan '02

I want Zope to run in the init.d as a daemon. Every time I've tried to set it up to run /zope/start it holds the process. It doesn't run in the background. I've tried copying code from the forum on Zope.org but no luck. Is there and easier way to make Zope an auto-run service.. this is import if we need to reboot the server. Thanks Allen May MCSi Web Admin 800.842.1702 x7628

2 1

Product->refresh broken in zope-2.4.3?
by Sven Rudolph 18 Jan '02

18 Jan '02

Hello all! After "refreshing" my product I can't create new instances of my product any more. I get the following error message: 'Object of type "None" is not callable' This happens even if I didn't change anything in my source code. Just clicking the "refresh" button is enough. When I restart the sever everything works fine again. This behaviour is new in zope-2.4.3 (I updated from 2.4.1) Is this a known bug? Must I downgrade again to 2.4.1? Thanks in advance Sven -- Sven Rudolph GermanMedicalServices.de GmbH Unter den Eichen 5, 65195 Wiesbaden Tel.: 06 11 / 97 46 25 2

2 1

Zope Not Being Logical
by Wade Pearce 18 Jan '02

18 Jan '02

Ok, Recently we had a site crash due to a corrupt Data.fs file. We attempted to recover the file using fsrecover.py but could never get it to run properly with Zope. We always got a Proxy Error: Connection Refused. So, we loaded a dated Data.fs file that is some 450MB smaller. Now, within zope/var, we have two Data.fs files: - Data.fs is the dated Data.fs at 245MB in size - Data.fs.Oz is the attempted recovered Data.fs at 695MB in size. But the strangest thing is happening. If we attempt to load Data.fs.Oz with Zope, it fails to work. But after attempting this, we restart Zope with the older Data.fs and the site contained within Data.fs.Oz appears. To further confuse matters, Data.fs.Oz continues to grow and when we create *.zexp files, they are coming from Data.fs.Oz. The older Data.fs file has not changed size since we discovered this back on January 10. Below is an ls -l of the zope/var directory over a period of time this morning: golum:/var/lib/zope/var# ls -l total 1738132 -rw-r--r-- 1 angel1 angel1 245093502 Jan 10 20:52 Data.fs -rw-r--r-- 1 root root 700849746 Jan 18 10:58 Data.fs.Oz golum:/var/lib/zope/var# ls -l total 1750088 -rw-r--r-- 1 angel1 angel1 245093502 Jan 10 20:52 Data.fs -rw-r--r-- 1 root root 707298826 Jan 18 11:22 Data.fs.Oz For some reason, Zope has loaded Data.fs.Oz and is using this. One thing I did notice was that when we tried to directly load Data.fs.Oz (as Data.fs), no z2-pcgi.pid file was created. Any ideas on what is causing this would be greatly appreciated. We've currently got a list of up-to-date *.zexp files from Data.fs.Oz, and during a quiter time, we're thinking of creating a clean Data.fs and uploading the data. The only problem we have is that we don't know if the clean Data.fs will inherit the problems of Data.fs.Oz because its buried within one of the folders. Cheers, Wade Pearce

2 1

apache + vhm & http-rfc compliance
by Oliver Bleutgen 18 Jan '02

18 Jan '02

I'm cc'ing one author of a vhm howto on this, couldn't find more contacts for relevant adresses. There is a small issue with the often used rewrite rules for apache + virtual host monster. All examples I have seen look like (sorry for the wrapping) RewriteRule ^/(.*) http://localhost:25080/zopefolder/VirtualHostBase/http/%{HTTP_HOST}/Virtual… [P] But, as I had to find out, this isn't completely HTTP/1.1 compliant, in so far as the client may send a host-header _with_ port number, i.e. Host: www.hostname.tld:80 which breaks above rules. When trying to leech my own site with wget, I was very astounded to get an 404 error, regardless what I entered as an URL. I use an extra condition to avoid that problem and do another rewrite if hostname has the port appended: RewriteCond %{HTTP_HOST} ^.*:80$ RewriteRule ^/(.*) http://localhost:25080/zopefolder/VirtualHostBase/http/%{HTTP_HOST}/Virtual… [P] http://localhost:25080/zopefolder/VirtualHostBase/http/%{HTTP_HOST}/Virtual… [P] Just wanted to mention that before internet explorer 7.8SP13 decides it wants to use the above rfc-option and a ton of zope-sites break seemingly for no known reason. Maybe VHM could/should be adapted? cheers, oliver

1 0

zope subtransactions with atomic commits
by Joseph Wayne Norton 18 Jan '02

18 Jan '02

Kapil - I have been doing some searching through some of the zope mail archives and the zope.org site. I ran across the following e-mail (quite old) from you ... http://zope.nipltd.com/public/lists/zope-archive.nsf/242bb7cc2b2c343d802568… I'm wondering if you have any further updates on the above subject. I have a similar requirement. I would like to be able to perform 1 or more "checkpoints" within a single zope request. If the request transaction is aborted, then only changes since the most recent "checkpoint" should be rolledback. Any ideas or suggestions? - j

2 1