No data in BAC DB (SQL) -- db_loader issue

I'm getting data in my dashbaord but no history. We have BAC 7.52 on Windows/SQL.

Here are the errors from the loader logs:

2009-11-30 08:10:14,393 [JMS Session Delivery Thread] (RecoveryPersisterAdapter.java:95) ERROR - RecoveryPersister.save(id, samples) returned collection of 1 samples - failed to save samples into Failure Recovery.

2009-11-30 08:10:14,393 [JMS Session Delivery Thread] (AbstractFailureState.java:29) ERROR - Failed to move chunk to Recovery
com.mercury.am.platform.dataentry.loader.core.LoaderException: Failed to save sample into Failure Recovery

2009-11-30 08:10:14,393 [JMS Session Delivery Thread] (RecoveryQueueWriterImpl.java:117) WARN - Faild to write samples to the queue trans_t.BACPROFDBO75.ftgprdbac404win.0
com.mercury.am.platform.dataentry.loader.persister.recovery.RecoveryQueueOverflowException: Unable to create new partition: maximum number reached (509 active partitions vs 509 maximum number allowed).

Any ideas?
  • Do you use MS SQL Server 2005 Enterprise Edition? If not, do you have Partition Manager enabled in BAC UI?
  • Thanks --- yes we have SQL Server 2005 Enterprise.
  • I am guessing you have data building up in the recovery queue in persist.dir and I would imagine your drive is begining to fill up.

    Did you recently turn on data purging on a BAC environment that has been online for a while? Are you getting any errors in the partition manager log on the DPS? There are a number of causes for this and none are an easy fix. I would run a dbverify to make sure everything is ok on the SQL side. Also I would have your SQL DBA take a look at things from a database perspective.

    I would contact support and have them help you to work through this issue. Uploading the dbverify log and loggrabber will help move through the support process.
  • Hi,
    First I'd like to point out that there're 3 different persistency mechanisms:
    1. GW: WDE -> DB Loader -> DB
    2. Online BLE -> DB
    3. Offline BLE -> DB

    Historic data isn't stored in DB points that either you didn't define default profile DB, or didn't marked CI/KPI's to be saved. There might be the 3d option - there's some problem with profile DB.
    The error in DB Loader regarding recovery queue specifies that the whole capacity of specific queue has been reached. Recovery queue isn't unlimited even if you have free space on disk. Look for other DB related errors around data insertion. Considering the fact that both DB Loader and historic data from BLE are not store there's some DB related problem. Check log/mercury_online_engine/Trinity.log for DB related errors as well. Do you use System health? It could point out what monitor is red.
  • Hi,

    I had the same problem and I observed that the size of the folder

    C:\HPBSM\.persist_dir\db_loader\recovery\ss_t.<bsm_profile>.<dbhostname>.0

    was of 7,95 GB. After we moved the files inside the folder in another disk and we restarted BSM (and we stopped Sitescope without restarting it). After the restart, in few minutes we had the same problem, now the folder is of 7,85 GB.

    In loader.log I have a lot of these errors:

     

    #########

    2012-08-27 11:38:40,839 [Main-Persistency-Reader-0] (MainReaderImpl.java:232) WARN  - Sample from main partition 469 was not delivered to SamplesConsumer (returned false). Partition will not be removed until next re-start:
    Sample_ss_t(BBSample, seq=3234430):
      eti_id                  = 'Binary[16]:0x67b72128d4674bbc8440ebba986331b2'
      u_iConnectionId         = '200165'
      szMeasurementName       = 'Telnet connections. Total opened'
      DowntimeState           = '0'
      u_iQuality              = '0'
      time_stamp              = '1.345650149E9' (2012-08-22 17:42:29)
      instance_id             = '1'
      ci_id                   = 'Binary[16]:0x341421405393845878a538b7461bdf9d'
      dTime                   = '1.345650149E9'
      profile_name            = 'SiteScope11'
      szMonitorName           = 'Connection Statistics Monitor'
      customer_name           = 'Default Client'
      u_iTargetId             = '200006'
      szTargetName            = 'sitescope_hostname.domain.it'
      ci_hint                 = 'SiteScope:22:200165'
      pmdb                    = '0'
      szCategoryName          = 'telnet connections. total opened'
      u_iSessionId            = '22'
      szConnectionName        = '10006'
      u_iCategoryId           = '200174'
      u_iMsmtQuality          = '0'
      u_iStatus               = '0'
      start_time              = '1.345650149E9'
      dValue                  = '0.0'
      u_iMonitorId            = '200166'
      measurement_description = 'Log Content Counter'
      szSessionName           = 'SiteScope11'
      u_iMeasurementId        = '200187'
      u_iHasProperty          = '0'

    2012-08-27 11:38:40,839 [Main-Persistency-Reader-0] (RecoveryQueueWriterImpl.java:157) WARN  - Failed to initialize partition for writing into the queue ss_t.bsm_profile.db1-vip.0
    com.mercury.am.platform.dataentry.loader.persister.recovery.RecoveryQueueOverflowException: Unable to create new partition: maximum number reached (509 active partitions vs 509 maximum number allowed).
     at com.mercury.am

    #####################

     

    Have you any idea of the problem?

    Thanks in advance

    Michele