Normal and reliable queue files
The key difference between disk queue files that employ the reliable(yes) option and not is the strategy they employ. Reliable disk queues guarantee that all the messages passing through them are written to disk first, and removed from the queue only after the destination has confirmed that the message has been successfully received. This prevents message loss, for example, due to syslog-ng PE crashes if the client and the destination server communicate using the Advanced Log Transfer Protocol (ALTP). Note that the
Both reliable and normal disk-buffers employ an in-memory output queue (set in quot-size()) and an in-memory overflow queue (set in mem-buf-size() for reliable disk-buffers, or mem-buf-length() for normal disk-buffers). The difference between reliable and normal disk-buffers is that when the reliable disk-buffer uses one of its in-memory queues, it also stores the message on the disk, whereas the normal disk-buffer stores the message only in memory. The normal disk-buffer only uses the disk if the in-memory output buffer is filled up completely. This approach has better performance (due to fewer disk I/O operations), but also carries the risk of losing a maximum of quot-size() plus mem-buf-length() number of messages in case of an unexpected power failure or application crash.
Size of the queue files
Disk queue files tend to grow. Each may take up to disk-buf-size() bytes on the disk. Due to the nature of reliable queue files, all the messages traversing the queue are written to disk, constantly increasing the size of the queue file.
The disk-buffer file's size should be considered as the configured disk-buf-size() at any point of time, even if it does not have messages in it. Truncating the disk-buffer file can slow down disk I/O operations, so syslog-ng PE does not always truncate the file when it would be possible (see the truncate-size-ratio() option). If a large disk-buffer file is not desirable, you should set the disk-buf-size() option to a smaller value.
|
|
Caution:
One Identity recommends that you do not build upon the current truncating logic of the disk-buffer files, because syslog-ng PE might pre-allocate the disk-buffer files and never truncate them in the future. |
NOTE: The disk-buffer file's size does not strictly correlate to the number of stored messages. If you want to get information about the disk-buffer, use dqtool (for more information, see Getting the status information of disk-buffer files).
NOTE: If a queue file becomes corrupt, syslog-ng PE starts a new one. This might lead to the queue files consuming more space in total than their maximal configured size and the number of configured queue files multiplied together.