Re: Help! my install of ONLINE 4.00 is 30% slower than Standard Engine

(((I tried to reply by email but it bounced for "bad system name beaudin"))) In article <855@beaudin.UUCP> john@beaudin.UUCP (John Beaudin) writes: > [SE versus OnLine benchmark details omitted -- OnLine is 25% slower] > Are there any comments from anybody as to why these results, assuming > ONLINE really is faster than SE, are valid? Maybe ONLINE isn't faster. I would be very surprised if Informix advertises anywhere that OnLine is >>always<< faster than SE in >>every<< configuration. You should keep in mind that OnLine is designed for a different use than SE, in particular for high-concurrency, high-reliability use. To test its strong points you should change your benchmark to: * spread the data over more than a measly half-meg table, which is probably mostly contained in buffers under either system, * throw 2, 10 or 20 concurrent processes against it, not just one. * include concurrent update processes with the queries, * take an archive without stopping the benchmark. Regarding your choice of query, "where description matches '*MEM*'" will always use a sequential scan. MATCHES (or LIKE) with a leading wildcard can't use an index. If this figures in your actual application you should consider redesigning the table so that the significant characters are initial letters of their column; then queries can use indexes to match them. Hope this helps, Dave Cortesi