tag:blogger.com,1999:blog-1755315926100938840.post477165369912471634..comments2024-03-28T15:16:06.294+03:00Comments on Erman Arslan's Oracle Blog: Exadata -- For Exadata Database Machine Admins -- Filtered InformationErman Arslanhttp://www.blogger.com/profile/07686231074805634082noreply@blogger.comBlogger32125tag:blogger.com,1999:blog-1755315926100938840.post-941578453517753982017-09-09T10:52:11.167+03:002017-09-09T10:52:11.167+03:00You need to use SNMP.
Even the Enterprise Manager ...You need to use SNMP.<br />Even the Enterprise Manager System Monitoring Plug-In (Plug-In for Power Distribution Unit) uses SNMP for monitoring the PDUs.<br /><br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-27405503607575822412017-09-09T06:56:48.815+03:002017-09-09T06:56:48.815+03:00Hi Erman,
Good post,is there any queries we can u...Hi Erman,<br /><br />Good post,is there any queries we can use to check PDU status?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-22265298968871820442016-10-18T12:53:22.215+03:002016-10-18T12:53:22.215+03:00Thank you for posting very detailed information or...Thank you for posting very detailed information <a href="http://www.lucidtechsystems.com/oracle-exadata-training/" rel="nofollow">oracle exadata</a>for exadata database machine.Nagendra Reddyhttps://www.blogger.com/profile/10578370690764479588noreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-8704840595913076252016-07-25T15:36:07.753+03:002016-07-25T15:36:07.753+03:00Thank you Erman
I will Try it First
Thank you Fo...Thank you Erman<br /><br />I will Try it First<br /><br />Thank you For your Reply..... :)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-18266224900001534752016-07-25T14:59:49.804+03:002016-07-25T14:59:49.804+03:00Check this Oracle Support note:
How to Replace a ...Check this Oracle Support note:<br /><br />How to Replace a 10G Ethernet Card on an Exadata Database Node (X3-2/X4-2) (Doc ID 1672412.1)<br />It is for replacing the ethernet card, but look what it says:<br />Pull out any SFP modules installed in the original 10G Ethernet card and install them into the replacment 10G Ethernet card making sure to orient them properly so that they seat fully into the card.<br /><br />So it seems it is pretty straigth forward :)<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-56823630888872619702016-07-25T14:43:08.404+03:002016-07-25T14:43:08.404+03:00Hi..Erman Arslan
Do you Have referece step by ste...Hi..Erman Arslan<br /><br />Do you Have referece step by step install SFP module on Oracle Exadata X3-2?? and how to configure on OS.<br /><br />Please your answer.<br /><br />Thanks<br />SyamsulAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-67257245741581253072015-06-26T02:09:23.220+03:002015-06-26T02:09:23.220+03:00Virtual Column filtering by default is not offload...Virtual Column filtering by default is not offloaded unless you set the paraemter "_CELL_OFFLOAD_VIRTUAL_COLUMNS". So in esence , smart scan doesn't offload virtual column filtering. Reference Exadata Smart Scan FAQ (Doc ID 1927934.1).<br /><br />Also refer to note "Find the statistics related to cell offload (Doc ID 1438173.1)" which states <br />"Full scan of a heap table will benefit from smart scan"<br /><br />Cheers<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-45170666930213148912015-06-25T08:43:54.338+03:002015-06-25T08:43:54.338+03:00Yes, it seems it does not matter..For example: In ...Yes, it seems it does not matter..For example: In doc :Exadata V2/X2-2/X2-8, X3-2/X3-8, X4-2 Customer Actions: Exadata Rack Down (Doc ID 1604373.1),<br /><br />It says;<br /><br />Prior to field engineer service actions:<br /><br />- Shutdown the Exadata Database Severs<br />- Shutdown the Exadata Storage Severs <br />- Powering Off Oracle Exadata Rack<br /><br />So, yes..Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-68504977899198482282015-06-25T02:37:24.900+03:002015-06-25T02:37:24.900+03:00"crsctl stop cluster -all on database server
..."crsctl stop cluster -all on database server<br />Then we need to shutdown Database servers First<br />Then shutdown Exadata storage servers<br />Lastly we need to power off the network switches and cut the power using power switches on PDUs."<br /><br /><<>>>><br /><br />Doc ID 1093890.1 is also correct and a practical way of doing it, as you prefer to SSH to cells from compute nodes and then carryout your tasks.<br /><br />I believe it doesn't matter as long as cluster services are shutdown properly.<br /><br />But normally exams questions follow the course ware.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-34212119085279815722015-06-01T19:48:48.859+03:002015-06-01T19:48:48.859+03:00What is your reference?
In Document : (Steps To Sh...What is your reference?<br />In Document : (Steps To Shutdown/Startup The Exadata & RDBMS Services and Cell/Compute Nodes On An Exadata Configuration. (Doc ID 1093890.1)) says ;<br /><br />9) Power off all cells:<br /><br /># dcli -g cell_group -l root poweroff<br /><br /><br />10) Copy the /opt/oracle.SupportTools/onecommand/dbs_group to /opt/oracle.SupportTools/onecommand/dbs_group2:<br /><br /># cp /opt/oracle.SupportTools/onecommand/dbs_group /opt/oracle.SupportTools/onecommand/dbs_group2<br /><br /><br />11) Remove the “first compute node name” from the /opt/oracle.SupportTools/onecommand/dbs_group2 file (using any editor, e.g. vi).<br /><br />12) Power off all the compute nodes (except compute node # 1):<br /><br /># dcli -g dbs_group2 -l root poweroff<br /><br /><br />13) Power off the compute node #1:<br /><br /># poweroff<br /><br /><br />14) Now, both Cell Servers & Compute nodes are down for maintenance.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-91168911596567534112015-06-01T19:40:37.927+03:002015-06-01T19:40:37.927+03:00correct sequence to properly shutdown the Exadata
...correct sequence to properly shutdown the Exadata<br /><br />crsctl stop cluster -all on database server <br />Then we need to shutdown Database servers First<br />Then shutdown Exadata storage servers <br />Lastly we need to power off the network switches and cut the power using power switches on PDUs.Anonymoushttps://www.blogger.com/profile/02371716225615427018noreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-47434579649702297892015-03-03T09:49:22.964+02:002015-03-03T09:49:22.964+02:00OPlan is a utility that facilitates the patch inst...OPlan is a utility that facilitates the patch installation process by providing you with step-by-step patching instructions specific to your environment for both in-place and out-of-place patch installation. Currently, OPlan is supported for Exadata Database Bundle Patches. For the latest information, see "Oracle Software Patching with OPLAN" in My Oracle Support Note 1306814.1Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-15489999580363494832015-03-01T23:54:37.890+02:002015-03-01T23:54:37.890+02:00Hi Sir,
Please, I am a bit confused about Oplan. ...Hi Sir, <br />Please, I am a bit confused about Oplan. I know Oplan is used to generate instructions specific to your environment. what about if you use RAC. Oplan will generate instruction for patching RAC oracle home and grid GRID infrastructure home or it will generate instructions for patching all the components on the database machine.<br /><br />Thank you.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-26763341244310790542015-02-16T14:52:10.913+02:002015-02-16T14:52:10.913+02:00great information ... thanks a lot Erman!!great information ... thanks a lot Erman!!Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-74729680162987565422015-02-09T15:32:13.485+02:002015-02-09T15:32:13.485+02:00The bonding and backup configruration is actually ...The bonding and backup configruration is actually It depends on your needs..<br />In the same paper there is another scenario.<br />Example Scenario 2: Gigabit Ethernet<br /><br />As you may see; there are some calculation there.. Throughput calculation.. According to these calculations:<br />For example: Customer wants 4 hrs to backup 6TB <br />So it makes @ 2TB/hr = 560MB/sec<br />According to this info, the configurations is done.<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-86673601769049086612015-02-09T15:26:34.588+02:002015-02-09T15:26:34.588+02:00Yes, bonding can be be used both for improving per...Yes, bonding can be be used both for improving performance and reliability.<br />In this post, I wrote "it improves reliability in Exadata", because in version X3, bonding was active-backup by default.<br /><br />When channel bonding is configured for the client access network during initial configuration, the Linux bonding module is configured for active-backup mode (mode=1). If a different bonding policy is preferred, then you may reconfigure the bonding module after initial configuration.,<br /><br />For example: In case of Infiniband; In the default deployment of the Exadata X4-2 Database Machine, the InfiniBand bonding is set up in the active/active mode, on all database servers and all storage cells. Earlier models of the Exadata Database Machine (V2, X2-2, X3-2) have the InfiniBand set up in the active/passive mode.<br /><br />(Oracle Exadata Database Machine - Backup & Recovery Sizing: Tape Backups : FIGURE 1 RMAN BACKUP USING INFINIBAND CONNECTIVITY AND ORACLE SECURE BACKUP)<br />Looking at the best practices, MAA makes Exadata compute nodes to connect to Media servers using Single Infiniband QDRs.<br />Each compute nodes uses 1 QDR to connect to IB switch. From IB switch to the Media servers, there are 2 QDR used per media servers. (active-active mode) <br />I didnt understand this one. Why using 2 QDRs between IB switch and media servers? If the answer is reliability then why dont they use multiple QDRs between computenodes and the IB switch.. Strange.<br />Anayways;continuing.. In MAA the Media Servers connect to the Tape Devices using 4 Fibers.<br />each Media servers connects to the tape drives using 4 fibers (in active active mode..) I think this is for increasing fibers throughput...<br /><br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-62870370896021576862015-02-09T14:50:56.572+02:002015-02-09T14:50:56.572+02:00Hi Erman,
Regarding bonding , I think the bonding...Hi Erman,<br /><br />Regarding bonding , I think the bonding improve the reliability and performance. i find the following: Network bonding enables two or more network interfaces to act as one, simultaneously, increasing network bandwidth and/or providing network interface redundancy.(Doc ID 738269.1), and the bonded ports can be configured for load balancing, fault tolerance and BETTER network utilization (Oracle Exadata Database Machine - Backup & Recovery Sizing: Tape Backups).<br /><br />Please tell me what do you think about it?<br /><br />Thanks - great blog.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-49972007203558972992015-01-13T12:25:31.735+02:002015-01-13T12:25:31.735+02:00I understand what you mean. Okay. But this blogpos...I understand what you mean. Okay. But this blogpost is not an exam review. The things written in here, is just some bunch of information which I have gathered while preparing the exam. The purpose of this blog post (also this whole blog) is to enhance te technical knowledge of mine and the ones who follow this blog. Thanks for you interest , really. I m glad for you to spend time reading and making comments about this post. The true and more understandable statement is then "allocate one RMAN channel to one tape drive." , "allocate two RMAN channel to two tape drives, one for each" and so on."Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-59758468870535111102015-01-13T11:43:25.661+02:002015-01-13T11:43:25.661+02:00I totally agree with all that you've just said...I totally agree with all that you've just said.<br />It's just a matter of how you understand of what you read.<br />I saw the simialr answer for the question:<br />'Which two are Oracle recommendations for media based backups performed for a database<br />running on a Database Machine?'<br />while preparing for the exam.<br />There is no word 'one rman channel' in this context, so equivalent number of channels and instances for me means<br />that number of channels should match number of instances (per tape drive). It is not said here that <br />one rman channel should be equivalent one tape drive.<br />Besides if your statement (answer) is correct there will be no option to choose another 100% correct answer for this question.<br />I think the problem is that answer is poorly formulated.Anonymoushttps://www.blogger.com/profile/01420847486031531147noreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-6053748913385934502015-01-12T18:00:04.573+02:002015-01-12T18:00:04.573+02:00So if you have two tape drives, again allocate one...So if you have two tape drives, again allocate one rman channel to one tape drive. Then you will have two rman channels mapped to these two tape drives..<br />I can not see why you find it not accurate?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-84999352145755199432015-01-12T17:57:51.816+02:002015-01-12T17:57:51.816+02:00Hi Timur,
Refencence: An Oracle White Paper Septe...Hi Timur,<br /><br />Refencence: An Oracle White Paper September 2011 Oracle Exadata Database Machine - Backup & Recovery Sizing: Tape Backups<br /><br />Maximum Availability Architecture (MAA) best practices recommend configuring the same number of <br />RMAN channels as the number of tape drives allocated for backing up the database. <br /><br />"In other words, allocate one RMAN channel to one tape drive. "Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-31808790082127233322015-01-12T16:48:15.917+02:002015-01-12T16:48:15.917+02:00Hi Erman,
official recommendation for tape-based b...Hi Erman,<br />official recommendation for tape-based baclup configurations goes like this:<br />Configure one RMAN channel per tape drive and add tape drives to scale backup rates<br />though this post is not accurate and may be misleading:<br />'For media based backups, it s recommended to allocate equivalent number of channels and instances per tape drive..' <br />Following this recommendation I can allocate more than 1 channel per tape which is contrary to BP:<br />http://www.oracle.com/technetwork/database/features/availability/maa-tech-wp-sundbm-backup-11202-183503.pdf<br />Do you agree?<br />Any references supporting your statement?<br /><br />Regards,<br />Timur<br />Oracle ACSAnonymoushttps://www.blogger.com/profile/01420847486031531147noreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-36421031182646320542015-01-12T08:35:36.839+02:002015-01-12T08:35:36.839+02:00Why not correct?
Reference :Migrating an Oracle E-...Why not correct?<br />Reference :Migrating an Oracle E-Business Suite Database to Oracle Exadata Database Machine (Doc ID 1133355.1)<br /><br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-22780911152150667972015-01-11T20:57:56.997+02:002015-01-11T20:57:56.997+02:00Hi Emran,
This is not correct in Exadata:
Transp...Hi Emran,<br /><br />This is not correct in Exadata:<br /><br />Transportable Database<br />-------------------------------------------------<br /><br />Source system must be 10gR2 (10.2.0.5) and little endian<br />Stricter service level requirements that obviate the required downtime with Oracle Data PumpAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-1755315926100938840.post-46021919037734350762014-11-22T21:28:32.543+02:002014-11-22T21:28:32.543+02:00What happened to your result ???What happened to your result ???Anonymousnoreply@blogger.com