Date: Fri, 29 Mar 2024 10:57:10 +0000 (UTC) Message-ID: <327994206.69921.1711709830128@ip-10-0-1-26.ec2.internal> Subject: Exported From Confluence MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_Part_69920_1409897730.1711709830127" ------=_Part_69920_1409897730.1711709830127 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Content-Location: file:///C:/exported.html
The File Activity panel displays = the top five database files with the highest relative activity since the la= st refresh. Available metrics allow you to view the top Reads/sec, Writes/s= ec, and Transfers/sec activity for all files or a selected database, disk, = or file. Unexpected spikes in the amount of file activity may warn you of i= neffective or excessive indexing or out-of-date statistics. For additional = information, see View file activity.
The <= strong>I/O Stall ms/sec chart displays the total time per = second (in milliseconds) that the SQL Server processes were waiting on= an I/O to complete. Chronically high I/O stall metrics may indicate a disk= bottleneck.
The <= strong>Reads/sec chart displays the number of physical rea= ds per second the disk sub-system is performing to bring SQL Server databas= e pages into the data (buffer) cache. If the reads rate is excessive for a = long period of time, consider taking action to avoid performance issues.
The <= strong>Reads KB/sec chart displays the size of the physica= l reads per second the disk sub-system is performing to bring SQL Server da= tabase pages into the data (buffer) cache.
The <= strong>Total KB/sec chart displays the total size of activ= ity per second across all databases on the monitored instance.
The <= strong>Writes/sec chart displays the number of physical da= tabase writes per second the disk sub-system is performing. If the writes r= ate is excessive for a long period of time, consider taking action to avoid= performance issues.
The <= strong>Writes KB/sec chart displays the size of the physic= al database writes per second the disk sub-system is performing.
The <= strong>Transfers/sec chart displays the total number of tr= ansfers per second across all databases on the monitored instance. Transfer= s are the sum of reads and writes. A higher transfer rate indicates a highe= r risk of resource issues.
There are circumstances under which the F= ile Activity panel does not display any data. For example, when you lose co= nnection to the Collection Service or SQL Diagnostic Manager Repository. An= other reason is the lack of collection data points. On new instances, file = activity collection occurs once per hour by default. By increasing the leng= th of time shown on the desktop client views, you may be able to see more d= ata points. You can modify the timeframe of data displayed in SQL Diagnosti= c Manager by clicking Tools >= ; Console Options while in the File Activity view and= changing the Keep data for the= last and Show= data for the last values.
You also can change the database collecti= on interval and grooming options to help you see more data points. Use the = Collect and alert on database m= etrics field in the General tab of the Monitored SQL Server Pro= perties window to manage your database collection. For grooming= options, use the Groom Session= s, Queries, Deadlocks, Waits, and History Browser data older than = field in the Grooming Options window.
Decreasing the collection interval and grooming your data less often can= result in performance issues and require additional storage space in your = Repository.
The File Activity pan= el displays results based on activity since the last refresh. You must have= two collections to create a point on the graph, three collections to creat= e a line. By default, the database collection interval is set at one hour, = which means that the File Activity panel does not show a point until an hou= r after adding the instance unless you perform a manual refresh. You can re= fresh and collect data by clicking the = Refresh button in the SQL Diagnostic Manager Manag= ement Console menu bar.
The History Browser a= llows you to view the state of your SQL Server instance at the time a = snapshot occurs, which is every six minutes by default. When using the Hist= ory Browser with the File Activity panel, note that data may appear in one = snapshot but not in others because no file activity occurred during that in= terval. Remember that file activity data does not appear for the first two = hours of monitoring a SQL Server instance because the metrics are base= d on the activity since the last database refresh.