RE: why archive vary in size

Well, one reason could be different amounts of activity hitting different parts of the disk while the archive is going on. Don't forget, at checkpoint time, the archive has to ensure that it has safely recorded any pages from the physical log which it hasn't yet archived so that it can preserve the database as it was when the archive began. Presumably, it grabs those pages into the archive at the point when it has to rescue them, and then has to record that it did so later on in the archiving process when it would normally have processed those pages (unless it manages to write the page to archive without needing to do more than record internally that it did so). I think blob updates or deletes (but not, normally, inserts) could also affect this - if people are updating blobs while the archive is running, it has to go and grab the before image from disk at checkpoint time since it isn't normally in the physical log. It will depend, in part, on whether the blobs are in a blob space or stored IN TABLE. Etc. Yours with an educated guess, Jonathan Leffler (jleffler@visa.com) #include <bother.ms-exchange.h> ---------- From: ANTS[SMTP:sui@naseej.com] Sent: Monday, April 06, 1998 11:04 AM To: informix-list@iiug.org Subject: why archive vary in size Hi, We are doing archive level 0 to disk daily. We noticed that the size of the file goes up and down greatly, more than the activities of the database itself. We have asked Informix, they dont have explanation. One guy said that it's because of logical logs that being backed up. So sometimes that logs are big and sometime they are small. But the Informix consultant said that archive level 0 does not save the logs in the level-0 only the log id. What's really going on??? We are using Informix 5.05 online. Nien =========================================== Mails are checked daily and processed right away. If you don't get my replies within 3 days please send your mail again.