CONTAP-46911：高負荷が原因で大規模なRDBの読み取りレイテンシが高くなることがある

最後の更新
PDFとして保存

Views:: 29

Visibility:: Customer

Votes:: 0

Category:: ontap-9

Specialty:: core

Last Updated:

問題

When a replicated database (RDB) instance experiences below, it can cause RDB read latencies to begin increasing over time.A volume of high concurrent read requests generated from external and/or internal workflows for an extended period of time(hours on end)
This is combined with frequent writes (transactions)
This can eventually lead to various management operations and commands timing out.
If this issue affects the volume location database (VLDB), it can affect data volume access causing NFS protocol delays and timeouts.
Additionally, such a high rate of concurrent read requests and ongoing transactions can result in the database journal file, rdb.db-wal, to grow large in size (that is, hundreds of megabytes or even several gigabytes).
In turn, this might lead to affected management processes going in and out of quorum, re-joining quorum taking extra time, and longer than normal process startup times.