I recently took over an ETL system which uses linked servers to pull data from a number of source databases, which I'm in the process of migrating over to SSIS. We have, on average, about 5 source databases per server & 5 warehouses on the other.Everything was fairly stable for years (I went through the log files to confirm this). On the odd occasion, they ran into this error:The operation could not be performed because OLE DB provider "SQLNCLI11" for linked server "???" was unable to begin a distributed transaction.The solution was to restart MSDTC & the problem would disappear for months. A few months ago, servers were upgraded to Windows Server 2012 & SQL Server 2014. Since then, we're basically being spammed by this issue.We've compared the DTC configuration of an old server (Win 2008, SQL 2008 R2) to the new 2012 server & everything looks the same. And it doesn't fail all the time. I'll say pretty much 50% of all ETL executions will report this error. A restart of this service only "fixes" it for a few hours. Another weird thing is that we have multiple ETL jobs on the same server. The linked servers for all these warehouses point to the same source servers. One ETL will run through fine, the other will fail.We've been through windows & SQL logs but can't pick up anything that might tell us why this is happening. Can't even replicate it on our test environment.Does anyone have any idea what I can try or where I can look to sort out this problem. Every site I go onto just goes on about the configuration. Can it even be related to something like memory or CPU usage?
↧