Hi

For me it is not clear, what can i do whith the module?
Think I can connect to a cassandra DB, but it is only an API?
Have you implemeted functions to outsource some drupal functions like sessions or caching?

A lot of questions.
Hope you can make it clear.

thx
bennos

Comments

jthomasbailey’s picture

you can test it out and tell us if it melts your site, heh heh

subscribing

waleedq’s picture

this module provides api for other drupal modules to communicate with cassandra's server you can check the readme included in the package also check the example module included in the package

tell now there is no support for sessions or caching through cassandra but you can implement them using the api from this module ,

paul kim consulting’s picture

For one, you can store all the watchdog events in Cassandra. It would be nice to build an abstraction layer between the Drupal API and the DB (via Thift?) or use the D7 data abstraction layer to write a DB plugin/driver. It wouldn't be able to do everything that the D7 DB layer needs (like JOINS), but if you could offload some data to Cassandra (for example, the node table?), that would be nice.

doublejosh’s picture

Status: Closed (fixed) » Active

This sounds like a thread I'd like to pay attention too.
Would be great to pick and choose what tables to offload to a no-schema db. Depending on your setup and circumstances some tables can get huge (especially if you're aggregating content) and some records rarely need to get accessed again after the first few times (like ubercart orders.) Here are some of the elements I imagine could potentially benefit...

  • stats and access logs - accesslog, history, simplenews_statistics (etc.)
  • nodes - node, node_revisions
  • appropriate fields - contnent_field_xxx
  • cache tables - cache_content
  • access permission - node_access, users_roles (content access is quite the performance hit.)
  • flags - flag_content, flag_counts
  • aliases and redirects - path_redirect, flag_counts
  • core profiles - profile_values (though I use content profile.)
  • aliases and redirects - path_redirect, url_alias
  • search - search_index, search_dataset, search_node_links, search_total (though SOLR is the right thing.)
  • sessions - sessions
  • taxonomy - term_data, term_node, term_synonym
  • commerce - uc_orders, uc_payment_receipts, uc_product_options (if you're lucky!)
  • voting - votingapi_vote (if you're lucky!)

Kinda of interesting that the thing I've always considered an Achilles heel of Drupal (one node table) might end up being the thing that makes it a highly scalable CMS by leveraging the abstraction layer and mixing no-schema goodness with SQL organization.

Update: Here is the resource about separating Drupal tables off... http://drupal.org/node/291373
I'm planning to do this as a first step, before I transition to a different DB technology for specific data.

doublejosh’s picture

Also, hey look at these...
http://drupal.org/project/mongodb
http://drupal.org/project/couchdb

I'm still a total noob with regard to this stuff, I just know I need to listen. Wonder if we'll see modules pop-up for: Hadoop, SimpleDB (AWS), Tokyo Cabinet or even BigTable (via App Engine.)

A little farther into future land is this: http://drupal.org/project/sparql_views

ClearXS’s picture

Strange that Hadoop is not implemented yet; its features look great, although I'm not into Apache. But it shows-up as I google on terms like: extended columns directory index file Apache table sorter; like to have more columns and sortable discriptions in http://all.drumi.org/
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

I go for the nuke, CIA, FBI, Bin Laden, Tsunami a.s.o. resistivity. Sometimes servers of free media are confiscated, so it would be nice being able to survive such a nuclear winter. Maybe I die, but then 'the FreeWorldWideWarrior.Cyb.org' will continue fighting the bad cyborgs!

I've heard that one can have a hidden server (Tor?), and one or two connected to that one; one with Boost and the other with Varnish + local sites that are mirrored on the hidden central site.

But this system goes way beyond? Feature request of-course to implement such a hidden server, local server, back-end, front-end, like different server where Drupal articles are placed and the published articles are read from.

Wikipedia:Tor_(anonymity_network)
TorProject: Hidden-Services
TorProject: Tor-Hidden Service
Onion-Router:Locating-Hidden-Servers.pdf

Wikipedia:Varnish_(software)
Drupal:Varnish - Varnish-Cache.org
KristianLyng:Varnish Backend Selection through DNS

Drupal:Boost - Handbook Boost

Proxies, accelerators, and more: OpenDNS with VPN or Tor ....
OpenDNS services: OpenDNS bypassed by XB Browser and OperaTor (Using Tor anonymity network)

http://en.wikipedia.org/wiki/Nginx ( nginx random_index for some nice other work )
"lightweight, high-performance Web server/reverse proxy and e-mail (IMAP/POP3) proxy, licensed under a BSD-like license."
http://en.wikipedia.org/wiki/Reverse_proxy
"proxy server that is installed on a server network or on network equipment. Typically, reverse proxies are used in front of Web servers. All connections coming from the Internet addressed to one of the Web servers are routed through the proxy server, which may either deal with the request itself or pass the request wholly or partially to the main web servers"
http://drupal.org/project/filefield_nginx_progress
http://github.com/masterzen/nginx-upload-progress-module

bennos’s picture

Views and Caching Panels and Panels Pane are missing.

Ravi.J’s picture

Status: Active » Closed (fixed)

There is now support for Varnish support for panels through ESI module

doublejosh’s picture

Status: Active » Closed (fixed)

Wanted to add http://redis.io to the list: http://drupal.org/project/redis (no releases, 5/11)