HDR recovery in asynchronous mode
Posted in 2009
Topics: High Availability & Replication, Performance & Tuning
Hi, If primary server crashes in an HDR pair (asynchronous mode say DRINTERVAL=5), How can we prevent the data loss if storage of primary server is accessible for read/write mode but IDS instance is unable to start. Can we apply the primary logs by some manual method or some other way on HDR-secondary (for transactions which are committed on primary node but was not sent on secondary node. Also approximately how much performance degradation can be faced for setting DRITERVAL from 30 to 5. Do we need to change the DRTIMEOUT as well? regards, Kamran
You can't. You can't. There is a cost for the 100% availability that = SYNC mode provides. That said, I can't see why there would be a significant impact in reduc= ing the DRINTERVAL down really low. = "KAMRAN HAQ" = <khaq@i2cinc.com> = Sent by: = To ids-bounces@iiug. ids@iiug.org = org = cc = Subj= ect 07/09/2009 10:40 HDR recovery in asynchronous mod= e AM [16289] = = = Please respond to = ids@iiug.org = = = Hi, If primary server crashes in an HDR pair (asynchronous mode say DRINTERVAL=3D5), How can we prevent the data loss if storage of primary server is access= ible for read/write mode but IDS instance is unable to start. Can we apply t= he primary logs by some manual method or some other way on HDR-secondary (= for transactions which are committed on primary node but was not sent on secondary node. Also approximately how much performance degradation can be faced = for setting DRITERVAL from 30 to 5. Do we need to change the DRTIMEOUT as w= ell? regards, Kamran ***********************************************************************= ******** Forum Note: Use "Reply" to post a response in the discussion forum. =
So can someone tell what is "dr.lostfound" for? Is there something related to "dr.lostfound" w.r.t. HDR recovery