May 24, 2013 at 12:05 am
Hi there!
Background: We're running log shipping between two servers, and the secondary server restores the transaction logs with standby mode, to support queries against the databases for reporting purposes.
After going through our perfmon logs and the disk metrics we, surprisingly, saw that the log disk on the secondary server, where we only have the .ldf files, for the standby databases had pretty high values for IOPS and the read/write ratio was about 95:5.
I would expect the disk where the data file exists to have more IOPS, and a relatively high value for reads (the queries from users, etc) and for writes (restoring the transaction logs as part of the log shipping job), but not really the log disk ...
How does it work, when SQL Server reads the transaction log file on the secondary databases, is it during restores to keep track of where the next transaction log should be applied? Can anyone please explain? I feel that there's a part missing in my log shipping puzzle. 🙂
Thank you very much!
Sincerely,
Gord
May 24, 2013 at 3:41 am
GordonLiddy (5/24/2013)
Background: We're running log shipping between two servers, and the secondary server restores the transaction logs with standby mode, to support queries against the databases for reporting purposes.
Where does the TUF files stored? Is it in same drive of ldf file?
May 24, 2013 at 5:09 am
They seem to be stored in the data files directory on the secondary server, e.g "E:\SQLDATAFILES\MSSQL10.INSTANCE1\MSSQL\DATA", where the .mdf files are placed.
That's not the same drive as where we keep the .ldf files. That drive has about 65% less IOPS than the log drive.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply