- Zope - Zope lists

RE: [Zope] Re: Caching and Proxy Servers
by Tiller, Michael (M.M.) 27 May '04

27 May '04

> From: zope-bounces(a)zope.org [mailto:zope-bounces@zope.org] On Behalf Of > Subject: Re: [Zope] Re: Caching and Proxy Servers > > On Thu, May 27, 2004 at 03:14:52PM +0200, Nate Aune wrote: > > > Here is another question for you. I notice in my profiling that two > > > things are just killing me. They are: > > > > > > here/getAllowedTypes > > > portal.list_filtered_actions(here) > > > > > > I have associated both of these with a RamCache that is not very > > > restrictive (in terms of REQUEST contents, i.e. AUTHENTICATED_USER, > > > URL1) and yet they never get cached. Any idea why? They are python > > > scripts (in case that matters). Any more clever ideas? > > Python scripts are usually fine candidates for RAM caching. > What do you mean by "they never get cached"? What do you see in > the cache statistics? > > What exactly *is* in your REQUEST variable settings for this > cache manager? I'd suspect something's not right there. RAM Cache Settings: Request Variables: 'AUTHENTICATED_USER', 'URL1' (each on their own line) Threshold entries: 1000 Max. age: 3600 Cleanup interval: 300 The "URL1" is in there so that it gets cached on a per URL basis. Is that the right way to accomplish that? It seems to work for everything else. Doesn't seem to work is defined as four hits to the same page (URL) by the same user gives the following in my statistics section: Path: /<plone_instance>/getAllowedTypes Hits: 0 Recent Hits: 0 Misses: 0 Memory: 0 Views: Entries: 0 Odd that it even mentions it?!? (with everything being zero) -- Mike

1 0

Help! Serious performance problem! (PTS? Zeo? SMP?)
by Jean-Francois.Doyon＠CCRS.NRCan.gc.ca 27 May '04

27 May '04

Hello, For reasons I can't seem to grasp, my Zope has gone real slow on me ... I switched from using Localizer + TranslationService to using the collective's PlacelessTranslationService (PTS) 1.0fork-rc7 ... And since then this occurs. On the good side, this seems to have fixed my leakage problem however! I've tried increasing the number of threads but it doesn't really seem to help, as well as the connection pool. CPU usage seems to be ~25-30% which isn't bad. (It's 4 PIII 700Mhz) I've tweaked the interval check as per some docs found on zope.org ... There's 4GB of RAM, which is plenty. I've set zeo-client-name, but that didn't seem to do much, so now I've set the "client" parameter in the zodb_db -> zeostorage section, and now the cache file DOES appear (Is this a bug? Shouldn't have zeo-client-name done the job?). Also is there anyway to pre-fetch/fill the caches in the threads? I'm wondering if the act of filling those caches might be what's slowing things down? In which case maybe I should try with less threads?? Of course a lot fo this started when I started using PTS ... Could it be the culprit somehow? I use Apache with mod_proxy ... Apparently the requests pile up faster than Zope can take them ... Sometimes I can stop Apache and watch Zope keep working for minutes after that (Sometimes longer than I'm willing to wait, at which point I shut it down!). Any thoughts/ideas/potential solutins would be most welcome! Thanks, Jean-François Doyon Internet Service Development and Systems Support / Spécialiste de dèveloppements internet et soutien technique Canada Centre for Remote Sensing / Centre Canadien de télédétection Natural Resources Canada / Ressources naturelles Canada http://atlas.gc.ca Tel./Tél. : (613) 992-4902 Fax: (613) 947-2410

2 1

ZCatalog keyword search unexpected results with operator=and
by Alex 27 May '04

27 May '04

I'm trying to use a ZCatalog for the first time and I'm getting unexpected results. Ultimately I'm trying to be able to search one or more keyword indexes, searching each one with one or more keywords, and be able to either AND or OR across the board, but I can't seem to be able to get AND working within an index. I am using a keyword index on a list field. I have manually updated the catalog and verified the index contents. Zope version is 2.7.0. In a Python script I have : results=context.catalog(categories={'query':['big','shiny'], 'operator':'and'}) If some objects are big, and some are shiny, but none are big and shiny, I get no results - as expected. But if some objects are big, and there are no objects that are shiny, then I get all objects that are big - I expected to get no results. Is this by design or am I doing something wrong ? Thanks, Alex

2 1

RE: [Zope] Piped Logs
by Bill Etheredge 27 May '04

27 May '04

[WME] Z2.log -- Apache can direct its access log (comparable to Z2.log) to a process rather than a file. Thanks - Bill ------------------------------------------------------------------------ --- On Thursday 27 May 2004 14:54, Bill Etheredge wrote: > Does zServer provide any capability comparable to Apache httpd's 'piped > logs' -- writing log files through a pipe to another process rather than > a file? syslog? -- Toby Dickenson

1 0

Zope locks up
by Bert Vanderbauwhede 27 May '04

27 May '04

Hi, I had a serious problem with my site yesterday. For some unknown reason, Zope seemed to lock up or slow down extremely, so we couldn't get a response anymore. I tried several things: * restart Zope * replace Data.fs by an earlier version * redirect Apache to another Zope server These things didn't work. After a couple of minutes the problem returned. After searching through Apache log files, a colleague found that something weird happened with requests for pages that started with '/error/'. So we changed the rewrite rules to deny those requests, and everything seemed to work again. Since then, we've only had the ususal (read LDAP) problems. My questions: * Were those '/error/'-requests really the cause of the problem? How can I be sure. And why did this happen just now? I've been using those same rewrite rules for over a year now, but I never had this problem before. * Is there a list of things to check for when Zope locks up? Yesterday, we looked at about everything, before we stumbled upon those '/error/'- requests. Could we have found this earlier? * Is there a list of bad requests, ie. requests that lock up Zope. I already denied WebDAV-requests (PROPFIND, LOCK) on Zope's HTTP-port, because of problems with M$ Word and Word-documents. Are there other things that I should deny, or at least handle with care? My configuration: Redhat 8.0 Apache 2.0.40 Zope 2.6.4 CMF 1.2 Bert Vanderbauwhede... -- "I laugh in the face of danger. Then I hide until it goes away." -- Alexander LaVelle Harris

3 5

Re: Caching and Proxy Servers
by Nate Aune 27 May '04

27 May '04

> I got caching of the navigation stuff working nicely. It isn't as > efficient as it could be because it constructs the "tree" for every URL, > but at least it doesn't re-render for the same URL thanks to caching. The nav_tree in general is not very efficient, and the sooner it gets rewritten, the better. This one is from Alan Runyan (runyaga) <tal:checkperms tal:define="global checkPermission nocall:here/portal_membership/checkPermission; global wtool nocall:here/portal_workflow;"> <metal:portlet use-macro="here/portlet_navigation/macros/portlet" /> </tal:checkperms> > Here is another question for you. I notice in my profiling that two > things are just killing me. They are: > > here/getAllowedTypes > portal.list_filtered_actions(here) > > I have associated both of these with a RamCache that is not very > restrictive (in terms of REQUEST contents, i.e. AUTHENTICATED_USER, > URL1) and yet they never get cached. Any idea why? They are python > scripts (in case that matters). Any more clever ideas? No, I don't know why they aren't getting cached. Maybe one of the caching gurus on the list can answer that one. Nate

2 1

conditional dtml looping
by J-Town Productions Ltd. 27 May '04

27 May '04

Can I use dtml-in to loop through objects with a certain value for a given property. Using dtml-in with an if statement such as <dtml-if "visible=='show'"> does not seem to work since the sequence still includes the items What would the syntax be for dtml-in if "visible=='show'">?

2 1

Virus gefunden in Nachricht "Dein Zeug's!"
by Dr. Christian Schütt 27 May '04

27 May '04

Symantec AntiVirus hat einen Virus in einem Anhang gefunden, den Sie (zope(a)zope.org <zope(a)zope.org>) an Mailing-Boxes(a)median-kliniken.de gesendet haben. Um sicherzustellen, daß die Empfänger die gesendeten Dateien verwenden können, führen Sie einen Virensuchlauf auf Ihrem Computer durch, entfernen Sie Viren aus infizierten Dateien, und senden Sie diesen Anhang erneut. Anhang: Tools.zip Virusname: W32.Sober.G@mm Durchgeführte Aktion: Säubern fehlgeschlagen : Isolierung erfolgreich : Dateistatus: Infiziert

1 0

Editing the modification_date
by KJZZ Webmaster 27 May '04

27 May '04

Is it possible to edit the bobobase_modification_date of an object using dtml or a python script? Sometimes I have to update an object that was entered into the Catalog say a month ago. However, afterward, when I search the Catalog and display the results according to date, I do not want this object to appear as if it a recent addition to the list. Thanks, John T.

2 1

New users can't log into the ZMI
by Steph Smith 27 May '04

27 May '04

I'm new to both Zope and Python (not to mention OO programming :), but slowly feeling my way around them :) I have a fresh install of Zope 2.6.4-r1, installed from ebuild. I've read through the "Installing and Starting Zope" chapter in the Zope Book, and also "Getting Started with Zope for Linux and Solaris" article on DevShed. After starting my zope instance, the "Getting Started" guide suggests that the first thing I do is log in to the ZMI as the init user, which I was able to do without any problem. The guide then suggests that I create a new user with manager rights, and to access the ZMI with this user going forward. I go into the acl_users object, and add a new user (with "Manager" role), which appears to be successful. However, when I log out as the init user, and try to log back in as my new user, I am unable to log back in to the ZMI. I have checked the password repeatedly, and that is not the problem. I've tried assigning the "Owner" role to my new user. Still no luck. Created other users. Again, no luck. Unfortunately, nothing is being logged in /var/log/zope/zope-instance, so I have no way of determining whether this is an authentication issue, or a permissions issue, or something else all together. I've looked under the Security tab of the acl_users object, and explicitly set permissions for Managers. I don't think I'm supposed to need to do this, and it didn't help anyway. I have made sure that the 'domain' field for the user has been left empty. Now I'm wondering if there's a step that I've missed, which must be done before I can start adding new users? Or if there is a way to increase what gets logged. I've looked inside the error_log object, but didn't find anything useful. Or perhaps somebody can offer another suggestion/direction to look? I'm very excited about zope (and its forthcoming staging ability). But when first starting out in new territory, its always these little things that trip one up! Thanks in advance! Steph Smith

2 1