On 12/12/2019 3:37 AM, Hongxu Ma wrote:
And I found my "full sync" log:
"IndexFetcher Total time taken for download
(fullCopy=true,bytesDownloaded=178161685180) : 4377 secs (40704063 bytes/sec) to
NIOFSDirectory@..."
A more question:
Form the log, looks it downloaded all segment files (178GB), it's very big and
took a long time.
Is it possible only download the segment file which contains the missing part?
No need all files, maybe it can save time?
I'm finding the code in the replication handler to be difficult to
follow ... but I do not think there is any way to have it do a partial
or catchup backup copy. I believe it must copy the entire index every time.
To achieve what you're after, using an alternate backup strategy would
be advisable. There are two good ways:
1) Assuming you're on an OS that supports them, make a hardlink copy of
the data directory to an alternate location on the same filesystem.
This should happen almost instantly. Then at your leisure, copy from
the hardlink copy to your final destination. When finished, delete the
hardlink copy.
2) Use rsync (probably with "-avH --delete" options) to copy from the
current index directory to another location. This will skip any
segments that already exist in the destination.
You could potentially even combine those two options in your final solution.
I do not think the replication handler deals with tlog files at all.
The transaction log capability did not exist when the replication
handler was built.
Thanks,
Shawn