- Zope - Zope lists

RE: [Psycopg] Zope crashing
by Mayers, Philip J 19 Sep '01

19 Sep '01

Hmm very puzzling. 0.99.6 - compiled as follows: ./configure --with-postgres-libraries=/usr/lib/pgsql --with-postgres-includes=/usr/include/pgsql --with-python=/usr/bin/python2.1 --with-mxdatetime-includes=../egenix-mx-base-2.0.2/mx/DateTime/mxDateTime BTW - to compile with that version of psycopg and the egenix package: diff -r psycopg-0.99.6/config.h.in psycopg-0.99.6-orig/config.h.in 4c4 < #define MXDATETIME_API_MODULE "mx.DateTime" --- > #define MXDATETIME_API_MODULE "mxDateTime" Python is: [root@hdb /tmp]# python2 Python 2.1.1 (#1, Aug 20 2001, 14:54:12) [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-85)] on linux2 Type "copyright", "credits" or "license" for more information. recompiled source RPM from the Python web site. Is it worth me running an --enable-devel version? What overhead is there? Regards, Phil +------------------------------------------+ | Phil Mayers | | Network & Infrastructure Group | | Information & Communication Technologies | | Imperial College | +------------------------------------------+ -----Original Message----- From: Federico Di Gregorio [mailto:fog@mixadlive.com] Sent: 19 September 2001 18:03 To: Mayers, Philip J Cc: 'psycopg(a)lists.initd.org'; 'zope(a)zope.org' Subject: Re: [Psycopg] Zope crashing On Wed, 2001-09-19 at 18:59, Mayers, Philip J wrote: > I'm still seeing periodic/sporadic crashes of my Zope installation, and it's > getting *REALLY* annoying :o) > > I sometimes get a core file (if I'm lucky) whose backtrace normally reads: > > Program terminated with signal 11, Segmentation fault. > <snip> > #0 0x401364fe in __select () from /lib/libc.so.6 > (gdb) bt > #0 0x401364fe in __select () from /lib/libc.so.6 > #1 0x4031708c in __DTOR_END__ () from /usr/lib/libpq.so.2 > #2 0x4030ecad in PQgetResult () from /usr/lib/libpq.so.2 > #3 0x4030ee1c in PQexec () from /usr/lib/libpq.so.2 > #4 0x402f9acb in curs_switch_autocommit () from > /usr/lib/python2.1/site-packages/psycopgmodule.so > #5 0x402fa0e8 in curs_switch_autocommit () from > /usr/lib/python2.1/site-packages/psycopgmodule.so this is _strange_ because ZPsycopgDA *does* *not* support autocommit and there is no way to enable it while using psycopg with Zope. [snip] > #32 0x8058aef in PyEval_CallObjectWithKeywords () > <etc down to 138> > > Note the depth of the call stack... noted. > Is this a problem with psycopg or Zope? Is anyone else seeing it? I'm unable > to repro these consistently even under high load, which leads me to think > there's an edge case coming into play here - but I don't know what. With -M > logging turned on and using requestprofiler, I see: sincerely i don't know. i am running some SQL Methods right now (with an --enble-devel compiled psycopg) and curs_switch_autocommit() does not get called. what version of zope/psycopg are you using? how have you compiled it? thank you for the report, federico -- Federico Di Gregorio MIXAD LIVE Chief of Research & Technology fog(a)mixadlive.com Debian GNU/Linux Developer & Italian Press Contact fog(a)debian.org 99.99999999999999999999% still isn't 100% but sometimes suffice. -- Me

1 0

ZCatalog/python script question
by ksmith＠99hats.com 19 Sep '01

19 Sep '01

Is something like this possible in python scripts? <dtml-in Catalog size=5> <dtml-var title> </dtml-in> I'm hoping to take advantage of Catalog's lazy results in batching. Thank you in advance, Kevin Smith

2 1

Newbie - Question to Gilles
by Toh Wanda 19 Sep '01

19 Sep '01

(or whoever's kind enough to answer) Hi! I (21, from Germany) 'm new to this list and quite a newbie (as far as zope is concerned) yet, so please be patient. My question: As you can create DMTL documents via a HTML-form (opposed to creating it directly in the management enviroment), I was wondering whether one can also create user accounts. My first tries don't work, instead of performing the action "acl_users/manage_users" you just get to the acl_user screen, without a new account. Well I never really expected this first shot to work anyway. Then I searched your list for a solution and found an advanced question by Gilles Lenfant, in which he describes exactly what I'm looking for as an issue already working. So, I was wondering Gilles, whether you would tell me, how you did it? (I didn't find a possibility to contact you directly. Isn't there any, or did I miss it?) That's about it, thanx in advance for any help, Tohwanda! _________________________________________________________________ Downloaden Sie MSN Explorer kostenlos unter http://explorer.msn.de/intl.asp

2 1

Problems with dtml-tree (again)
by Juan García García 19 Sep '01

19 Sep '01

Hi, the tag dtml-tree creates a table in HTML, but i want to put a header(Quantity) to this table. Do you know any way to do it? Example: ----+--------------------+----------- | | | Quantity | +---+--------------------+----------+ | - | Item 1 | 9 | +---+--------------------+----------+ | | - | Item 1.1 | 5 | +---+---+---+------------+----------+ | | | + | Item 1.1.1 | 4 | ----+---+---+------------+----------- Bye.

2 1

RESPONSE.redirect(to a subanchor#everything behind disappears)
by Hannes Wyss 19 Sep '01

19 Sep '01

Hi everybody For a form-evaluation i want to redirect the user back to the form in case there is input missing. this is what I try: <pre> <dtml-call "RESPONSE.redirect('some_url?foo=some&bar=vars#subanchor')"> </pre> and this is what i get: some_url?foo=some&bar=vars annoying! I have concluded (knowing very little about Python), that somewhere within the Python-Part of the redirect method the tail of my url is interpreted as a comment. The Python documentation states, that # is indeed the start of a comment, unless it is part of a string literal. I would greatly appreciate any clues how to circumvent this problem... thanks in advance! -- mit freundlichen Grüssen / best regards Hannes Wyss Konzeption & Entwicklung +41 1 350 85 86 www.ywesee.com > intellectual capital connected > www.nachahmer.ch

2 1

Re: [Zope] variable SQL statements - from external methods
by ed colmar 19 Sep '01

19 Sep '01

Hi All. Jim and several others were helping me to get my variable SQL methods working right. They're great and do everything I want form the test window. When I call them from my external method, they come back with a REQUEST error. This is how I call it: mastertablename = "sometablename" self.SQL_create_master_table(mastertablename=mastertablename) Then I get: Error Type: Bad Request Error Value: ['mastertablename'] Whats the syntax to shove this variable and it's value inside the request? Or am I doing something else wrong? Thanks everyone! -ed- > CREATE TABLE <dtml-var mastertablename> > ( > .... > )

4 3

Re: Zope Developer Recruiting Site was Great Zope Site Idea?
by lozinski＠openstepnews.com 19 Sep '01

19 Sep '01

Thank you for your posting recommending a zope site listing resumes and jobs. In general this would be a great thing to do, but sadly there is little market demand for it. I speak as a recruiter who develops his site in Zope. www.jobmart.com. I currently am the dominant, indeed only remaining recruiter for Apple's WebObjects developers. I have seriously considered becoming a zope recruiter, and growing my web site to provide such services, but sadly I believe that there is little or no market for such services. Sure developers would love to post their resumes there, but hiring managers are easily able to find developers, and so would not be willing to pay for such a site. So it is tough to justify the investment in building such a site. The other problem is the general lack of demand for web services now. Investment is just not flowing into the web, so there is excess of zope developers. Regards Chris 1-510-795-6086 lozinski(a)openstepnews.com lozinski(a)jobmart.com

1 0

RE: [Zope] Question: Virtual Site Hosting on Zope
by sean.upton＠uniontrib.com 19 Sep '01

19 Sep '01

Unfortunately, I haven't written a howto, but I'm working on doing something along those lines for internal purposes at my company. I'll likely get around to doing a howto eventually. I hope you don't mind that I posted this back to the list, but I figured it would be useful to folks... A few tips: - Run Squid on a dual-homed box, with a public and private network interface. Your web servers / ZEO clients would be on the private interface for security reasons. - Redirectors simply rewrite URLs so that Squid can get content from somewhere other than the intended URL. pyredir (you can find it on freshmeat) is a good redirector for rewriting URLs in squid, written in python, and uses a simple syntax and regexes for a config file. It has to be hacked to support things like http POST and HEAD requests (by default, it only rewrites GET requests). I also hacked pyredir to make logging optional, since I didn't see much value in keeping a rewrite/redirector log (I didn't have log rotation on this, and ran into a 2GB file-size limit, so I just made it optional). I redirect to nodes.foo.com, for example, which is set up to have several IP addresses associated with it in /etc/hosts. You need to make sure that the URL you are redirecting to has the correct URL for Virtual Host Monster. - You will be setting up Squid in "accelerator mode" which is different than normal proxy, in that it accepts normal http requests, unlike a proxy server. It caches pages, which is good, unless you do not want them cached. Make sure you set up ACLs in squid.conf to prevent the caching of ZMI pages using the no_cache directive. Also, if you want, bind Zope management interface access with another Squid ACL to the private interface, so it can only be accessed from the private network; this works out well if your changes are only made from a protected office network or the like. - Supporting multiple ZEO client nodes in squid involves 2 things: compiling Squid with "external DNS" support, which means you use an external daemon to do lookups... in newer versions of Squid, instead of doing this, Squid uses internal lookup code instead of your machine's resolver libraries, which prevents correct round-robin behavior. Compile with external DNS support and make sure you spawn enough external 'dnsserver' daemons to handle the lookup load... Keep in mind that in an accelerator, this needn't be that high, because squid will cache the IP address lookups after the first time. You will need to create multiple entries in /etc/hosts with the same domain name; when the squid dnsserver process queries for a lookup, it will return to Squid multiple IP addresses. Squid will internally round robin them equally weighted (at least if the box is under anything above a moderate traffic load). In order to support this, you will likely need to use a source install of Squid (for example, the binary Debian package for Squid isn't compiled with external dns support). - Supporting Virtual Hosts in Zope with Squid is the same as in Apache, you just use the squid redirector instead of mod_rewrite rules, and you use a Virtual Host Monster. I hope this is useful info. Sean -----Original Message----- From: Steve Drees [mailto:drees@rangebroadband.com] Sent: Tuesday, September 18, 2001 12:33 PM To: sean.upton(a)uniontrib.com Subject: RE: [Zope] Question: Virtual Site Hosting on Zope any chance you've written a how-to up on this? :) > -----Original Message----- > From: zope-admin(a)zope.org [mailto:zope-admin@zope.org]On Behalf Of > sean.upton(a)uniontrib.com > Sent: Tuesday, September 18, 2001 12:17 PM > To: mike(a)flipper.eskimo.net; zope(a)zope.org > Subject: RE: [Zope] Question: Virtual Site Hosting on Zope > > > Consider using Squid as an http accelerator, a Squid redirector > (internally > rewrites URLs prior to proxy; I use pyredir - it is written in Python, and > is easy to modify), and a virtual host monster. Squid can load balance > multiple back-end web servers (well, round-robin), which makes this an > especially useful setup if you ever see youself using ZEO. This setup > bypasses the need for using Apache to interface with Zope, and Apache > running on a different port can be proxied to as well by squid, based upon > how you have set up your redirector rules. > > Sean > > -----Original Message----- > From: Michael Murphy [mailto:mike@flipper.eskimo.net] > Sent: Monday, September 17, 2001 10:17 AM > To: zope(a)zope.org > Subject: [Zope] Question: Virtual Site Hosting on Zope > > > I know this will be an easy question for this mailing list... Is there > generally considered to be a 'best way' of doing Virtual Site Hosting > on Zope? I have skimmed some HOWTOs on www.zope.org, but I am afraid > the material may be a bit dated and a 'better way' may be generally > known. Any help would be appreciated! > > Mike > > > _______________________________________________ > Zope maillist - Zope(a)zope.org > http://lists.zope.org/mailman/listinfo/zope > ** No cross posts or HTML encoding! ** > (Related lists - > http://lists.zope.org/mailman/listinfo/zope-announce > http://lists.zope.org/mailman/listinfo/zope-dev ) > > _______________________________________________ > Zope maillist - Zope(a)zope.org > http://lists.zope.org/mailman/listinfo/zope > ** No cross posts or HTML encoding! ** > (Related lists - > http://lists.zope.org/mailman/listinfo/zope-announce > http://lists.zope.org/mailman/listinfo/zope-dev ) >

1 0

MySQLDA woes with 2.4.1 RH6.1
by Richard Moon 19 Sep '01

19 Sep '01

MySQL is the most popular database on the web. <RANT> Is it really not possible to include the MySQLDA with the main Zope distribution ? I'm trying to use Zope to meet clients' business needs and losing money working out why I'm getting the following message when running setup.py (why I have to do it I don't know but after you unpack the MySQLDA-python distributiion and read the README you find you have to do it !) ((No I don't want to know why I have to do it )). I'm assuming I have to have the source Zope distribution to get the missing Python.h for Python 2.1 ? Aaagh. Great job you guys doing all this Open Source stuff but soooo frustrating that you can't use it without being a C and Python wizard. It's not like its some obscure, little used feature I'm after here. I remember that survey on the Zope site that said connection to RDBMs was the number one interest of Zope users. How many people have given up on Zope for this kind of reason - thousands and thousands I would guess. Most customers don't complain, they just go elsewhere !!! </RANT> Anyway its looking for Python.h running build_ext building '_mysql' extension gcc -g -O2 -Wall -Wstrict-prototypes -fPIC -I/usr/include/mysql -I/opt/Zope-2.4o _mysql.c:31: Python.h: No such file or directory _mysql.c:40: structmember.h: No such file or directory error: command 'gcc' failed with exit status 1 The only Python.h I've got is in the Python 1.5.2. that comes with RH6.1. I was deliberately running the Python 2.1 that comes with Zope (as it tell you to), but presumably that's why I get the problem ? Thanks in advance Richard Moon richard(a)dcs.co.uk

1 0

memory leaks on python21 and zope240
by Paul Zwarts 19 Sep '01

19 Sep '01

Greetings, I wish I could make this a detailed message showing that I actually know what I'm doing but that isnt possible. I have a few zopes installed from 233 on py152 to zope 240 on py211. The python211 versions get a HUGE memory leak.. I have no idea what is causing it. It started happening I think when I recompiled python (it went fine with no errors). Can anyone make a suggestion? This is Redhat 7.1 Thanks, Paul Zwarts

1 0