Calling Acp Neeku Rac Interview Pedutunna..tondaraga Raa!

[littlestar]

[quote name='Silence..Please' timestamp='1324399325' post='1301170682']

ipudu morethan 75% of clients have RAC .. critical application aithe .. 90% RAC untundi
[/quote]

RAC kanyte mundu DB lo clustering concept lekunde na?

[Achilles_afdb]

[Achilles_afdb]

db lo antha chaduvukuna vaale

[littlestar]

[quote name='Silence..Please' timestamp='1324399650' post='1301170725']
yup .. each node will have its own vip .. but still .. u r accessing the same data source using different ip's .. u do that because.. if failover happens .. ur weblogic server can still connect to other nodes which are still accessible eventhough one node is down
[/quote]

yup ..meeku paina app server unte ...maaku webserver/load balancer untadi

[Gachibowli_Diwakar]

[quote name='AFDB SREERAAMUDU' timestamp='1324399616' post='1301170721']

aa bongu endi configure chesesi..maaku emina issue vaste shut abort..ante.vere chance istalledu..inka endi nerchukundi..

inka twist enti ante shut abort work kakapote host reboot ..[img]http://www.desigifs.com/sites/default/files/brahmam_style2_0.gif?1290348889[/img]
[/quote]
shutdown abort .. work avvakunda undadhu .. one thing u can do is.. go and kill Oracle processes at host level using "kill -9 " command .. LOCAL=YES ani kanipinche oracle sessios ni kill chesey

shutdown abort chesina .. kill -9 tho kill chesina okate .. its not at all a clean shutdown ..

DB upgrade cheyyaali annapudu .. when u shutdown the db using "shutdown abort" make sure u bring up the db .. and shutdown it cleanly using "shutdown immediate" and then do any upgrade .. ledhu ante naasanam ye inka

[Spartan]

[quote name='Little Star' timestamp='1324399826' post='1301170741']

yup ..meeku paina app server unte ...maaku webserver/load balancer untadi
[/quote]

[img]http://img530.imageshack.us/img530/6233/venkyyeah.gif[/img]

[PMREDDY19]

[quote name='Little Star' timestamp='1324399187' post='1301170664']
PMR ki clustering gurinchi chepthunnava
next reply 10 lines pakka
[/quote]
[quote name='Silence..Please' timestamp='1324399239' post='1301170673']
naku application side aithe no idea bhayya.. kani as far as I saw .. JDBC thin client use chesi connect authuu untaaru db's ki .. i think that is via weblogic servers ..

inka db ki ela connect avvochu ante .. nuvu unna server meedha db TNS entry unte .. Oracle binaries install ayi unte aah server meedha .. using sqlplus u can connect ..
[/quote]






Exactly this is the way we conenct to DB.

but we used some kind of embedded url than a RAC bcoz of issues we faced with Oracle DB RAC cluster in my previous project more than an year ago..

but now in the current set of priojects we are refreshing, we had a big chunk of DB RAC clusters comign up..
got to see how I can Deal with this Oracle RAC cluster this time..




edho okati cheyyali iagaa..

[littlestar]

[quote name='PMR aka OM' timestamp='1324399898' post='1301170746']






Exactly this is the way we conenct to DB.

but we used some kind of embedded url than a RAC bcoz of issues we faced with Oracle DB RAC cluster in my previous project more than an year ago..

but now in the current set of priojects we are refreshing, we had a big chunk of DB RAC clusters comign up..
got to see how I can Deal with this Oracle RAC cluster this time..




edho okati cheyyali iagaa..
[/quote]
[url="http://www.oracle.com/technetwork/middleware/weblogic/oraclewls-rac-1-131306.pdf"]http://www.oracle.com/technetwork/middleware/weblogic/oraclewls-rac-1-131306.pdf[/url]


idi saduvko..gud info

[Gachibowli_Diwakar]

[quote name='Little Star' timestamp='1324399705' post='1301170731']

RAC kanyte mundu DB lo clustering concept lekunde na?
[/quote]

undhi .. kani adhi at host level .. like for example .. HP provided some clustering mechanism .. peru gurthu ledu ..

kani aah cluster failover mechanism valla side effect enti ante .. for example 2 hosts ki clustering chesaav anuko .. oka host meedha services run authunapudu .. for some reason host crash aithe .. aah clustering mechanism failover node meedha start ayi .. volumes ni mount chesi .. services ni up chesthundi .. so u still didnot achieved that 100% availability

with RAC u can achieve that ,.. barring not all nodes gets crashed

[PMREDDY19]

[quote name='Little Star' timestamp='1324399826' post='1301170741']

yup ..meeku paina app server unte ...maaku webserver/load balancer untadi
[/quote]

ippudu nenu cehsetidhi antha gidheynaayya,, danekka no weblogic project moving forward to aplication deppoyment level, okaa naa ko kuda inthaku mundu weblogic lo ledu ikkada.

migtrating all of them towards weblogic now..


inka e Apache ki ithey ..

First DNS -->> GLB ----> VIP ( Local Load Balancer) ---> Apache Webservers>>


inka idhi kaani External customer facing sites ithey ,,

Prathi Dagagra bokkalo boley firewalls untayi.. aa appp servers emo De-Militiraized Zone(DMZ) layers lo untayiii..


motham apache antha,, e path ni securre gaa unchadamey..



Classes cheppesukovachu emi nenu inka ilagey..

[Gachibowli_Diwakar]

[quote name='AFDB SREERAAMUDU' timestamp='1324399683' post='1301170728']

so how scan in RAC differes fromVIP.both are used for load balancing and failover.
[/quote]

naku no idea ba SCAN IP gurinchi .. 11g r2 nunchi ochindi idhi .. 2months lo I willbe configuring RAC .. apudu cheptha

[PMREDDY19]

[quote name='Little Star' timestamp='1324400041' post='1301170758']
[url="http://www.oracle.com/technetwork/middleware/weblogic/oraclewls-rac-1-131306.pdf"]http://www.oracle.co...ac-1-131306.pdf[/url]


idi saduvko..gud info
[/quote]

hmm inka sadukonudu enti vayya.. real time lo edho oka server meeda prayogallu sesudey direct gaa..

oka ti Dobbesthey ,rendodhi careful gaa sestamu..

ayina book save chei pettina future lo refer cheyyaniki..

[littlestar]

[quote name='Silence..Please' timestamp='1324400048' post='1301170759']

undhi .. kani adhi at host level .. like for example .. HP provided some clustering mechanism .. peru gurthu ledu ..

kani aah cluster failover mechanism valla side effect enti ante .. for example 2 hosts ki clustering chesaav anuko .. oka host meedha services run authunapudu .. for some reason host crash aithe .. aah clustering mechanism failover node meedha start ayi .. volumes ni mount chesi .. services ni up chesthundi .. so u still didnot achieved that 100% availability

with RAC u can achieve that ,.. barring not all nodes gets crashed
[/quote]

[Spartan]

[quote name='Silence..Please' timestamp='1324400048' post='1301170759']

undhi .. kani adhi at host level .. like for example .. HP provided some clustering mechanism .. peru gurthu ledu ..

kani aah cluster failover mechanism valla side effect enti ante .. for example 2 hosts ki clustering chesaav anuko .. oka host meedha services run authunapudu .. for some reason host crash aithe .. aah clustering mechanism failover node meedha start ayi .. volumes ni mount chesi .. services ni up chesthundi .. so u still didnot achieved that 100% availability

with RAC u can achieve that ,.. barring not all nodes gets crashed
[/quote]

yup...RAC uses FAN mechanism...a failover times lo....ade advantage....so as to notifyother processes about configuration and service information such as service up/down events....etc....

[littlestar]

[quote name='PMR aka OM' timestamp='1324400187' post='1301170772']

hmm inka sadukonudu enti vayya.. real time lo edho oka server meeda prayogallu sesudey direct gaa..

oka ti Dobbesthey ,rendodhi careful gaa sestamu..

ayina book save chei pettina future lo refer cheyyaniki..
[/quote]screenshots tho untadi..datasources config vi..