So far we're thinking it might be
related to examine doing internal indexing. The searching works in general, so
it seems as if the indexing is succeeding. During the downtime, proc mon shows
lots of activity at the following paths, which clears up exactly when the
process becomes responsive again.
which we've got defined in our
ExamineIndex.config file, which we've more or less left alone.
We're running a load balanced
environment with the media files stored on a network share. The examine indexes
are specific to each web server.
We see a lot of "[UmbracoExamine]
Cannot index queue items, another indexing operation is currently in
progress," in the log, but that isn't any more prevalent when we're down
than under normal operation.
Has anyone else experienced this
sort of behavior? Is there a way to prevent index building from blocking end user requests?
Umbraco locks up for about 3 minutes about once a day per server
So far we're thinking it might be related to examine doing internal indexing. The searching works in general, so it seems as if the indexing is succeeding. During the downtime, proc mon shows lots of activity at the following paths, which clears up exactly when the process becomes responsive again.
[SITE]/App_Data/TEMP/ExamineIndexes/Internal and
[SITE]/App_Data/TEMP/ExamineIndexes/InternalMembers
which we've got defined in our ExamineIndex.config file, which we've more or less left alone.
We're running a load balanced environment with the media files stored on a network share. The examine indexes are specific to each web server.
We see a lot of "[UmbracoExamine] Cannot index queue items, another indexing operation is currently in progress," in the log, but that isn't any more prevalent when we're down than under normal operation.
Has anyone else experienced this sort of behavior? Is there a way to prevent index building from blocking end user requests?
is working on a reply...