Hi support community,
I'm facing an issue on a CRS 5.0(1) with historical reports (HDS).
Last known issue on the Windows Server was that there were no free space on hard drive.
This is now fixed and everything looks fine again.
Now the user reporting that they could not get historical reports.
As far as I could troube shot this issue by now, I can say that all regarding services are up and running.
I can still run reports for the current day but not for the last week or month.
So I guess there is a issue that the CDR don't stay in the MSSQL server.
I go through the logs and could not find anything suspicious.
Just one thing gives me to think of, in the database "da_cra" / table "ConfigLog":
- last entry was on 07.11.2012 where the space issues started
- I could not get any usefull information about the last entries
CRS 5.0(1) is not supported by TAC anymore so I hope some one could put me in to the right direction for further trouble shooting.
The user need the HDS reports until a new soloutin for CC is online.
My action plan looks as following:
- enable tracing for SQL - HDS
- perform check table / repair table on MSSQL
Unfortunately there is a message when I try to enable the tracing, but I don't get it what this message try to say me.
I don't know any MSSQL commands to run such checks / repairs.
Did you perform purging on this database already ? If not, please try doing that and purge the oldest data so that some free is left for the new entries.
Is this a High Avaliability setup?
Pls rate helpful posts !!
this is not a HA environment.
Do I have to run purging manual? If so how do I perform it?
As far as I could see it from the logs and configuration, auto purging is running every day at midnight.
Thanks and Regards, Alexander
Purging Historical Data
As the Cisco CRS Engine runs, it collects information about the status and
performance of the CRS system. Historical information is stored in a database that
can then be accessed to provide reports (see the Cisco Customer Response
Solutions Historical Reports User Guide for more information about Cisco CRS
When the database approaches its maximum size, some or all of the data in it must
be removed. Removing data from a database is called purging.
When the system purges data, it removes data from the db_cra database. It
determines what information to purge based on the number of months you specify
and on the current date. For example, if you instruct the system to purge data older
than 12 months, a purge on January 15 will purge data older than January 15 of
the previous year.
Note When you purge data, you permanently delete it. If you want to keep data that will
be purged, back up the database.
Please refer the section "Purging Historical Data" from page 354 in the below link for more information.
Hope this helps.
Please rate helpful posts by clicking on the stars below the right answers !!
let me see if this could help.
I'm facing an issue that we could not get data older than current day.
So if the database would be reached the maxium size so I guess we could not get the latest CDR but older one, right?
But in our situtation we have a vice versa situation, we get current CDR but not historical.
I checked the purge settings:
- schedule is configured to daily "12:00 AM"
- purge data older than "99" months
- notify when database reach "70" % of 2048 MB
- initiate autmoatic purge when databse size exceeds "80" % of 2048 MB
- auto purge data for the oldest "15" days
So this should be fine
I attached the PurgeProcess.log and have some questions about it:
line 4: Calling execute with purge date sat nov 27 00:00:03 CET 2004
--> where is this date coming from? date time settings are correct on Windows host
line 17: oldest record in the database has a timestamp of tue feb 26 07:27:39 CET 2013
--> this is excatly what I experience, any how the purge process gone crazy and deleting to much CDR's
line 18: database size exceeded auto-purge threshold of 80%
--> why is the database have such a great size if the lates record is from day before (call amount ~ 500 calls per day)
So at this point I have two questions why is the time stamp wrong and how could I clean up the database (what are these data which fill 80% of database)
Thanks and Regards, Alexander