could somebody please advice with following:
We have IPCC 7.0 designed as fully redundant and I want to achieve also HDS redundancy (duplication). I setted up two AWs in one site (one primary and one secondary) and enabled HDS on both AWs. Primary AW has central controller side A as preffered and secondary side B. Is this configuration correct? Is there any way how to synchronize both HDS databases (if they ware not created in same time)? Both AWs also act as WebView servers, is it possible to point WebView to local HDS?
Thank you for response, but I ment if it is possible to synchronize two HDS databases directly. There could be a case when you set second HDS longer time later than the first one and in such case some data are already purged from central controller database. When second HDS will finish synchronisation with central controller database than two HDS databases will be out of sync.
Yes, I tried to use ICMDBA but it can synchronize only central controller databases. But my main problem is with the design of HDSes. Some documentation say that you can configure also secondary distributor in site as HDS and some documentation says that only primary distributor should be configured as HDS. Also WebView documentation say that it should not be installed on secondary distributor.
Do you have both AWs in same admin site? Are both WebView servers reading data from local HDS? In my case it is also working but WebView on secondary distributor is taking data from HDS on primary distributor.
I also have the same question. How many HDS can you have, one or two. If two, how does replication work? Reading ICM 7.0 documentation and it does not really answer how many HDSs can be installed.
Cisco doesn't specify the number of HDS per logger, I made two HDS that sync with the same logger without problem. It also depends on the data size.
The data replication only happen between logger and HDS, not two HDS. Since the duplexed loggers are sync'ed each other, the data in two HDS can be seen as 'identical'.
Thanks for explanation. But those of you who run duplexed HDS what do you do in following situation. Let's say there is a 24/7 call center. HDS 1 is down for maintenance (or any other reasons). Few hours later it comes back online. Now HDS 1 database is different from HDS 2. Do you leave databases the way they are or do you manually sync them? Thanks.
In this scenario you do not need to manualy synchoronize two HDSs. When HDS 1 will be back online it will get missing data from logger. Problem is if data are already purged from the logger (I think that default is 14 days). In this case you will need to manualy synchronize HDS 1 against HDS 2. This could be done only using SQL tools (Microsoft SQL Server Enterprise Manager). ICMDBA tool can not synchronize one HDS with other HDS.
This is always a problem, Cisco doesn't recommend to use non-ICM tool to sync DB, because logger will check recovery key of HDS before replication, if this HDS is replicated by third-party tool, the recovery key may not match, and logger may not continue sync'ing with this HDS.
Hello, I opened TAC case regarding this issue some time ago and following is solution that was provided by TAC:
The procedure for rebuilding HDS2 database from HDS1 database.
1.) Backup both the HDS1 and HDS2 databases using SQL Enterprise Manager
2.) Shutdown ICM services on HDS2
3.) Using SQL Enterprise Manager delete the HDS2 database.
4.) Using SQL Enterprise Manager allocate a new HDS2 database the same size
as HDS1 database.
5.) Using SQL Enterprise Manager restore the HDS1 backup to the new HDS2
6.) Truncate the Recovery and Logger_Admin tables in HDS2 database.
7.) Restart ICM Services.