I work at VoltDB. There isn't currently a feature like this in VoltDB, for example like DBMS_JOB in Oracle.
You could certainly use a cron job on one of the servers in your cluster, or on some other server within your network that could invoke sqlcmd to run a script or echo individual SQL statements or execute procedure commands through sqlcmd to the database. Making cron jobs highly available is a general problem. You might find these other discussions helpful:
You could also look into something like rcron.
One thing to be careful of when converting from an RDBMS to VoltDB is that VoltDB is optimized for processing many small transactions in parallel across many partitions. While the architecture of serialized execution per partition excels for many operational and streaming workloads, it is not designed to perform bulk operations on many rows at a time, especially transactions that need to perform writes on many rows that may be in different partitions within one transaction.
If you have a periodic job that does something like "process all the new rows that meet some criteria" you may find this transaction is slow and every time it runs it could delay other parts of the workload, especially if many rows have accumulated. It would be more the "VoltDB Way" to replace a simple INSERT statement that you may be using to ingest data (to be processed later by a scheduled job) with a procedure that inserts and immediately processes the row of data. You might even need a procedure that checks for other records and processes small sets of rows as a group, for example stitching together segments of data that go together but may have arrived out of order. By operating on fewer records at a time within one partition at a time, this type of procedure would be more scalable and would keep the data closer to your desired finished state in real time, rather than always having some data waiting to be processed.
