Configuration of Downstream database for goldengate capture process in integrated mode

I have multiple source databases which are sending the archive logs to our Downstream mining database using Redo transport configuration.We have setup the goldengate extract on Downstream database which reads the archivelog and generate the trail file.I am facing an issue during peak hours when there are frequent redo log switches on the source database. The downstream database does't receive the archive log in sequence order.Few archive logs arrive first and few later but none is missing on downstream DB

The integrated extract Process on downstream database hangs when ever the archivelog arrive out of sequence and Logminer is not able to transition to mine the archive log. When I restart the extract it again starts mining the archive logcorrectly and lag starts decreasing. Below is the alert log message.Please help and suggest a solution.