I recently had to move my Oracle 11g Database from a Windows 2012 R2 server to Windows 2022 server and I've been getting intermittent "Oracle Communication: Failed to connect to server or parse connection string" errors. If I reboot the server, it goes away but comes back after some time. To migrate the database, I installed the Oracle software using the default setting for a database and altered the database to mirror what I had set in the old database, then I exported the data from the old database and imported it to the new database, no problem. I got all my data and PL/SQL code (packages, procedures, and functions) and all my apps access it just fine. Eventually I will be upgrading from 11g to 23AI, but for the purposes of migration I stuck with 11g to make things easier/quicker. In any event, I getting this intermittent errors logging into the database and I'm assuming I missed a setting somewhere. If I enter the username/password a few times (3 in the most recent case) I am eventually able to log in, but this issue is popping up during data processing via my apps and is causing problems with data. Anyone have any idea what is going on? I assume I missed a setting somewhere but have no idea what. The old database was up for 8 years on Windows 2012 R2 and never had this issue.
Official Oracle docs say that for a machine with more than 16GiB of RAM we need to allocate 16GiB of swap.
Our servers are RHEL 7 and have 256GiB of RAM.
DBAs do not want to see the system swap, so they want us to monitor the 16GiB of swap very aggressively.
I suggested that we double the RAM to 512GiB (the expense is approved), and disable the swap. However, this is against Oracle recommendation of having 16GiB of swap, even though we double the RAM.
Honestly, I don't get it how having 3% of swap makes any sense, or why if I am adding more RAM than we had swap, we have to keep the swap.
So, are there any good arguments I can use to justify running Oracle without swap?
P.S. The only reason I mention the doubling of RAM is to demonstrate the ridiculousness of the argument I am having a hard time arguing. What I am really looking for is arguments to justify disabling of the swap.
Our Oracle database(11g) is very slow all of a sudden. I donno where to start the debugging process to find what actually is the problem?
I tried generating table stats and looked for locks..but i can't find any information that points me directly to the cause, or maybe i'm looking at the wrong place..
Please guide me on where to start..
Well, I'll define what's Slow :)
Inserts take hell lot of time..even a 100,000 record insert takes abt an hour
Is there an easy way to determine to which RAC node of an Oracle 11g R2 system I am connected? I am trying to perform some failover tests and I want to make sure my application is correctly connected to one node and upon the shutdown of this node node makes the transition smoothly to another node without any noticeable delay on the front end. Maybe it is worth mentioning that we make use of TAF.
I considered using Enterprise Manager for this, but I guess that when I am connected to one node running em and this node goes down I will not really have a chance to monitor the nodes connectivity status.
I've recently received the dump file of the soon to be migrated Oracle Database. I've installed Oracle 10g and now I need to restore the original database.
Can anyone give me a step by step procedure for how to restore it from the *.dmp file ?