How to get conditions data from Frontier cache

Complete: 3



  • CMSSW version : CMSSW_1_0_0 or higher

Job Configuration

check frontier configuration for your local site

Check the calib-data section of ${CMS_PATH}/SITECONF/local/JobConfig/site-local-config.xml
How to config the calib-data section of site-local-config.xml is described here.

set up PoolDBESSource module with the following parameters:

  • string connect = "frontier://cms_conditions_data/detectorschema"
    Beginning in CMSSW_1_3_0_pre2, "cms_conditions_data" selects the default database but can be replaced with a frontier database name. Beginning with CMSSW_1_7_2 the oracle databases names behind frontier changed. These are the current and former frontier and oracle database name mappings:
    • FrontierProd (formerly Frontier) -> cms_orcoff_prod (formerly cms_orcoff)
    • FrontierInt -> cms_orcoff_int (formerly int9r_lb)
    • FrontierPrep (formerly FrontierDev) -> cms_orcoff_prep (formerly cms_orcoff_int2r)
    • FrontierOnProd -> cms_orcon_prod (only accessible from the .CMS network)
    • FrontierOn -> orcon (old ORCON hardware)
      If you need others send email to Dave Dykstra <>. The default frontier database is selected from site-local-config.xml and is currently set to FrontierInt at all sites.
  • prior to CMSSW_1_7_0_pre4 needed "untracked bool siteLocalConfig = true" which caused the parameters from site-local-config.xml to be used and a default "catalog" to be defined. If "siteLocalConfig" was not set to true, the extra parameters were not included and an "untracked string catalog" had to be added. Beginning in CMSSW_1_7_0_pre4, neither "siteLocalConfig" nor "catalog" are necessary any more (if they are present, they are ignored), and the calib-data parameters from site-local-config.xml are always included when the connect string is of the above form. Beginning with the same release, you can add additional frontier parameters in parentheses (parameters detailed here) after the database selector, for example:
         string connect="frontier://FrontierPrep(retrieve-ziplevel=0)/CMS_COND_FRONTIER" 
    Note - you can find the list of available detector schemas here.



Example configuration file ecal_pedestals_frontier.cfg for CMSSW_1_7_0_pre6 and later:
 process TEST = {
     es_source = PoolDBESSource {
         VPSet toGet = {
                 string record = "EcalPedestalsRcd"
                 string tag = "EcalPedestals_test17x"
         string connect =  "frontier://FrontierPrep/CMS_COND_FRONTIER"
         string timetype = "runnumber"
         PSet DBParameters = {
             untracked int32 messageLevel=0
     untracked PSet maxEvents = {
         untracked int32 input = 100
     source = EmptySource {
         untracked uint32 firstRun = 1
         untracked uint32 numberEventsInRun = 1 
     module get = EventSetupRecordDataGetter { 
         VPSet toGet = {
                 string record = "EcalPedestalsRcd"
                 vstring data = {"EcalPedestals"}
         untracked bool verbose = true 
     module print = AsciiOutputModule { }
     path p = { get }
     endpath ep = { print }
In older releases you may see some errors about CORAL_AUTH_USER not set; either ignore these messages or set both $CORAL_AUTH_USER and $CORAL_AUTH_PASSWORD environment variables to empty.
Prior to CMSSW_1_3_0_pre2, replace the "PSet DBParameters" section above with just:
     untracked uint32 messagelevel = 0 
Prior to CMSSW_1_5_X, replace the "PSet maxEvents" with the following in the EmptySource:
     untracked uint32 maxEvents = 100 

Run the example with this:

 cmsRun ecal_pedestals_frontier.cfg 



If you want to see Frontier debug and logging information, set the following environment variables:

Currently, writing the log file to an AFS file system is very slow and can greatly slow down cmsRun. You may want to direct it to a local disk such as somewhere under /tmp instead.

If there is a problem with a frontier cache server or a central frontier server, it is not likely to hang for long because the client will time out within 10 seconds of no activity and try another server in its list until the list is exhausted. You can enable debugging and tail the log to find out if it is in the middle of an operation.

There is also a script to extract just the SQL queries from the log file. It is not currently installed in CMSSW, it is in ~dwd/bin/frontierqueries on If a line begins with '*', that means it is a "short-lived" query, and if it begins with "-" that means that it was found in the local client cache and was not sent to any server. At the end of each line is a parenthesized pair of numbers like (MMM/NN%); MMM is the full size of the result of the query in bytes, and NN is the compression percentage (only present if compression is enabled).

Review Status

Reviewer/Editor and date Comments
ZhenXie - 09 Aug 2006 page author
JennyWilliams - 03 Apr 2007 moved page into swguide

Responsible: ZhenXie
Last reviewed by: Reviewer

Edit | Attach | Watch | Print version | History: r29 | r26 < r25 < r24 < r23 | Backlinks | Raw View | Raw edit | More topic actions...
Topic revision: r24 - 2008-05-05 - DavidFutyan

    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    CMSPublic All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback