Hi Stefan,
Am 28.10.2011 14:49, schrieb Stefan Küng:
> The status cache crawls folders as soon as it detects a change. This can
> lead to a heavy disk load if a folder is used a lot, for example if you
> run a build.
> To avoid such situations, the status cache process has a few safeguards
> in place:
> * Once a folder is crawled and its status fetched, that same folder
> won't get crawled again for about a minute.
> * when TSVN does an update/commit/..., it sends the cache a message
> telling it to block that folder and all folders below it. As soon as the
> TSVN operation is finished (i.e., the dialog showing the operation
> progress is closed), it sends the cache another message to unblock the
> folder again. The cache *then* crawls the folder and refreshes the status.
> * in case the cache encounters an error while crawling a folder, that
> folder is blocked from crawling again for two minutes. Such crawl errors
> can happen if the working copy database is currently in use (e.g.,
> another SVN client accesses it, a virus scanner has the file open
> exclusively, ...).
>
Thanks for the detailed description. :-)
Is something of this kind included in the manual?
If I understand it correctly, it seems that blocking a folder from being
crawled may last for a maximum of 2 minutes. If I add the maximum
crawling time (depending on the size of my working copy), I get the
maximum time I've to wait shortly after my last edit (if it caused an
error, which I cannot exclude nor surely detect)? ...and I don't know
how long a full crawl takes... this does not feel so good. :-(
Is there a way to explicitly re-trigger the crawling?
Best,
Wolf
------------------------------------------------------
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&dsMessageId=2868834
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].
Received on 2011-10-31 09:22:32 CET