toasters

toasters@lists.teaparty.net

3 participants
13514 discussions

RE: cluster using F740 and F760
by Traitel, Eyal 10 Mar '01

10 Mar '01

Todd, I cannot answer on behalf of our marketing group etc., but I would see a few obstacles: 1. Partner node needs to take care of both nodes activity in takeover mode, so it cannot be that much lower than other node. 2. We sell active-active clustering, so both nodes are active, and customer is not wasting his money for any passive protection. It leads you to the point that it may not worth the complexity to sell such a setup. 3. Technically, you must have same NVRAM size, and I believe you should also have reasonably close response times for any NVRAM mirroring and general underlying OS activity. That just makes non-equal nodes business more complex than one might think. 4. Anyway - customers reactions are always welcome here, and I'll run this to cluster devel. group, and see if they have to say anything.... Yours, Eyal. ---------------------------------------------------------------------- eTraitel - I'm the new eBuzzword around !!! ---------------------------------------------------------------------- Eyal Traitel - Filer Escalation Engineer CNA, MCSE, CSA, NetApp CA Network Appliance BV Holland Office Center Kruisweg 799b 2132 NG, Hoofddorp The Netherlands Office: +31 23 567 9685 Cellular: +31 6 5497 2568 Email: eyal(a)netapp.com ---------------------------------------------------------------------- Get answers NOW! - NetApp On the Web - http://now.netapp.com ---------------------------------------------------------------------- -----Original Message----- From: Todd C. Merrill [mailto:tmerrill@mathworks.com] Sent: Friday, March 09, 2001 6:11 PM To: Traitel, Eyal Cc: 'max chan'; toasters(a)mathworks.com Subject: RE: cluster using F740 and F760 From: max chan [mailto:maxchan0@yahoo.com] > Is it possible to use F740 and F760 as a cluster pair? On Mon, 5 Mar 2001, Traitel, Eyal wrote: > No, sorry. > This is unsupported. Ah, but is it *possible*? ;) You can put 1 GB of memory into an F740 to match the 1 GB in an F740. That is unsupported too, but possible. Ask me how I know. So, what's the real reason why it's not supported, clustering two non-exact-but-in-the-same-family filers? Is there something technically preventing it? Is is a marketing/sales decision? Is it to protect the unwashed masses from their own stupidity? Why not allow a lower-end filer take over a higher-end one? You're in a degraded mode in the first place during a failover condition, so why not allow a calculated risk? Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

3 2

RE: Multiple Oracle DBs on a filer
by Jennifer Armenta 09 Mar '01

09 Mar '01

Since putting everything on the filer would be ideal it's not a reality as you have stated just in case there is a problem. Im consolidating 6 unix servers running Oracle dbase to 840 filer. I was thinking it would be nice to run the ORACLE_HOME dirs and logs from 2 unix servers rather than 6. Wonder if anyone has thought about consolidating that way or even if it a good idea... -----Original Message----- From: Yarmas, Tom [mailto:Tom.Yarmas@netapp.com] Sent: Friday, March 09, 2001 1:14 PM To: 'Jennifer Armenta'; 'George Kahler'; toasters(a)mathworks.com Subject: RE: Multiple Oracle DBs on a filer > Is there a "best" method for where everything should be for > multi dbases? "Best" method, hmmm. I guess the easy answer is everything on the Filer, and if you need to save some space, have multiple instances share the ORACLE_HOME. But in reality, there are enough variables that it depends on what you are trying to acomplish and how much local storage your database machine has. Typically, I find ORACLE_HOME on the database machine storage, and all of the datafiles, redo logs, etc on the Filer. Often the boot drives are large and mirrored, so there is plenty of space. Plenty being relative ;) Of course this tends to preclude failover in the case of a failed database machine unless a backup machine has an ORACLE_HOME installed on it. And if you duplicate the ORACLE_HOME onto multiple machines for failover, then there is a risk that some change was not propogated to all machines. Of course, having multiple database instances within a single Filer volume has implications on SnapShots and SnapRestore/SnapMirror since these technologies work at the volume level. If you needed to do a SnapRestore for one instance, any others in the same volume would be affected. Similarly, when you take a SnapShot, the SnapShot would be for the entire volume, so it may be worthwhile to have multiple instances down (for a cold backup) or in hot backup mode for a hot backup. You generally do not want to have a seperate volume for each database instance unless they are large databases simply because each volume must have at least one parity disk, and you don't want volumes made up of a small number of drives if you can help it for performance. The only things that need to be on local storage are the alert logs and the trace files directory. That is so that if a connection to the Filer is lost, Oracle can write it's alert info for troubleshooting. If you can't talk to your Filer, you want to be able to see info in the alert log to take action. And if you can't talk to your local disks, you know immediatly. -tom

1 0

RE: Multiple Oracle DBs on a filer
by Yarmas, Tom 09 Mar '01

09 Mar '01

> Is there a "best" method for where everything should be for > multi dbases? "Best" method, hmmm. I guess the easy answer is everything on the Filer, and if you need to save some space, have multiple instances share the ORACLE_HOME. But in reality, there are enough variables that it depends on what you are trying to acomplish and how much local storage your database machine has. Typically, I find ORACLE_HOME on the database machine storage, and all of the datafiles, redo logs, etc on the Filer. Often the boot drives are large and mirrored, so there is plenty of space. Plenty being relative ;) Of course this tends to preclude failover in the case of a failed database machine unless a backup machine has an ORACLE_HOME installed on it. And if you duplicate the ORACLE_HOME onto multiple machines for failover, then there is a risk that some change was not propogated to all machines. Of course, having multiple database instances within a single Filer volume has implications on SnapShots and SnapRestore/SnapMirror since these technologies work at the volume level. If you needed to do a SnapRestore for one instance, any others in the same volume would be affected. Similarly, when you take a SnapShot, the SnapShot would be for the entire volume, so it may be worthwhile to have multiple instances down (for a cold backup) or in hot backup mode for a hot backup. You generally do not want to have a seperate volume for each database instance unless they are large databases simply because each volume must have at least one parity disk, and you don't want volumes made up of a small number of drives if you can help it for performance. The only things that need to be on local storage are the alert logs and the trace files directory. That is so that if a connection to the Filer is lost, Oracle can write it's alert info for troubleshooting. If you can't talk to your Filer, you want to be able to see info in the alert log to take action. And if you can't talk to your local disks, you know immediatly. -tom

1 0

RE: Multiple Oracle DBs on a filer
by Jennifer Armenta 09 Mar '01

09 Mar '01

Is there a "best" method for where everything should be for multi dbases? Oracle home on separate server or everything on the filer (database and Oracle home)? -----Original Message----- From: Yarmas, Tom [mailto:Tom.Yarmas@netapp.com] Sent: Friday, March 09, 2001 11:56 AM To: 'George Kahler'; toasters(a)mathworks.com Subject: RE: Multiple Oracle DBs on a filer There are 2 things to keep in mind here. First, there is the oracle account's home directory, and second there is an environment variable known as ORACLE_HOME. These 2 things can be different. ORACLE_HOME used to have to be unique among instances of Oracle, but with OFA in version 8.x, that has been somewhat relaxed. In versions prior to 8.x, it is more difficult for Oracle instances to share the ORACLE_HOME directory (but not impossible, at least not with unix). As a standard practice, I never install the Oracle software in the oracle account's home directory. That way I can put things in the home directory that affect multiple instances (i.e. shell scripts or sql scripts, etc.), and if I have multiple ORACLE_HOME directories then I don't have to copy things around when I make changes. Oracle will look at the ORACLE_HOME and ORACLE_SID environment variables to determine where it should find certain files by default. I have had various Oracle instances share the same ORACLE_HOME directory without consequence. One of the drawbacks of multiple instances sharing the ORACLE_HOME is that if you upgrade from one version of Oracle to another, it affects all of the instances that were using that ORACLE_HOME. That may not be a good outcome. I prefer that each Oracle instance have it's own ORACLE_HOME, if possible. You could of course create multiple oracle user account (i.e. oraprod1, oraprod2, or something more informative) instead of simply oracle. There is nothing magical in the user oracle and the group dba. They happen to be the default. However, in order to use different user names, you must re-link the Oracle S/W since the group is bound in to the oracle kernel at link time (trivial procedure). However, if multiple users are sharing the Oracle executables, then they must share the same unix group id (since it is linked into the code). Typically, I find that it is best to create a single oracle account (or a small number of oracle accounts for related instances), and use a shell script to display a choice of various Oracle instance names to the user. Then depending on the response, the ORACLE_HOME, ORACLE_SID and other variables are set. Then when the user logs in, all the environment variables are set correctly, and the ORACLE_HOME directory is different than the oracle account's home directory. And, if I install a new version of Oracle, it lives in it's own ORACLE_HOME directory, and each individual instance is not affected by changes made to another, I simply have to edit the script that lives in the oracle account's home directory so it sets the environment variables correctly. You should be able to create qtrees and/or volumes and mount points so that the Filer matches exactly what your DBA's see today in an existing environment, if that would make them happier. -tom > -----Original Message----- > From: George Kahler [mailto:george@YorkU.CA] > Sent: Thursday, March 08, 2001 10:05 AM > To: toasters(a)mathworks.com > Subject: Multiple Oracle DBs on a filer > > > Hi, > I'm setting up an 840 filer for our DBA group to run about > 4-to-5 Oracle > OLT databases and 4-to-5 Oracle DSS databases being served from about > 1/2 a dozen Unix servers. This is something new to the DBAs > in that they > are not experienced with NFS or anything like a filer; a common disk > storage. Their databases are married to specific Unix servers. The > project mandates that the environment needs to be set up in such a way > that if any Unix server fails, the database service that it was > providing could easily be brought up on any other available Unix > server. (So far no problem) > > I have created and mapped out a quota tree into which I installed the > Oracle distribution. It looks something like: > /softwr/oracle/product/7.3.4 > /7.3.5 > this is then exported to all of the Unix servers. > > The problem I have with is the Unix Oracle home directory that I also > created in /softwr/oracle/home, ie. > oracle:x:600:211:Oracle user:/softwr/oracle/home:/bin/ksh > but the DBAs are telling me that the oracle home directory > needs to be unique > for each Unix server and Database running on that server because the > home directory is the place where they keep their environment. > > How do other people deal with this? Do you create specific Unix > accounts for each Oracle database running on a server? > > Thanks > George > > -------------------------------------------------------------- > ----------------- > George Kahler e-mail: > george(a)yorku.ca > Sr. Systems Administrator humans: (416) > 736-2100 x.22699 > Computing and Network Services machines: (416) 736-5830 > Ontario, Canada, M3J-1P3 >

1 0

RE: Multiple Oracle DBs on a filer
by Yarmas, Tom 09 Mar '01

09 Mar '01

There are 2 things to keep in mind here. First, there is the oracle account's home directory, and second there is an environment variable known as ORACLE_HOME. These 2 things can be different. ORACLE_HOME used to have to be unique among instances of Oracle, but with OFA in version 8.x, that has been somewhat relaxed. In versions prior to 8.x, it is more difficult for Oracle instances to share the ORACLE_HOME directory (but not impossible, at least not with unix). As a standard practice, I never install the Oracle software in the oracle account's home directory. That way I can put things in the home directory that affect multiple instances (i.e. shell scripts or sql scripts, etc.), and if I have multiple ORACLE_HOME directories then I don't have to copy things around when I make changes. Oracle will look at the ORACLE_HOME and ORACLE_SID environment variables to determine where it should find certain files by default. I have had various Oracle instances share the same ORACLE_HOME directory without consequence. One of the drawbacks of multiple instances sharing the ORACLE_HOME is that if you upgrade from one version of Oracle to another, it affects all of the instances that were using that ORACLE_HOME. That may not be a good outcome. I prefer that each Oracle instance have it's own ORACLE_HOME, if possible. You could of course create multiple oracle user account (i.e. oraprod1, oraprod2, or something more informative) instead of simply oracle. There is nothing magical in the user oracle and the group dba. They happen to be the default. However, in order to use different user names, you must re-link the Oracle S/W since the group is bound in to the oracle kernel at link time (trivial procedure). However, if multiple users are sharing the Oracle executables, then they must share the same unix group id (since it is linked into the code). Typically, I find that it is best to create a single oracle account (or a small number of oracle accounts for related instances), and use a shell script to display a choice of various Oracle instance names to the user. Then depending on the response, the ORACLE_HOME, ORACLE_SID and other variables are set. Then when the user logs in, all the environment variables are set correctly, and the ORACLE_HOME directory is different than the oracle account's home directory. And, if I install a new version of Oracle, it lives in it's own ORACLE_HOME directory, and each individual instance is not affected by changes made to another, I simply have to edit the script that lives in the oracle account's home directory so it sets the environment variables correctly. You should be able to create qtrees and/or volumes and mount points so that the Filer matches exactly what your DBA's see today in an existing environment, if that would make them happier. -tom > -----Original Message----- > From: George Kahler [mailto:george@YorkU.CA] > Sent: Thursday, March 08, 2001 10:05 AM > To: toasters(a)mathworks.com > Subject: Multiple Oracle DBs on a filer > > > Hi, > I'm setting up an 840 filer for our DBA group to run about > 4-to-5 Oracle > OLT databases and 4-to-5 Oracle DSS databases being served from about > 1/2 a dozen Unix servers. This is something new to the DBAs > in that they > are not experienced with NFS or anything like a filer; a common disk > storage. Their databases are married to specific Unix servers. The > project mandates that the environment needs to be set up in such a way > that if any Unix server fails, the database service that it was > providing could easily be brought up on any other available Unix > server. (So far no problem) > > I have created and mapped out a quota tree into which I installed the > Oracle distribution. It looks something like: > /softwr/oracle/product/7.3.4 > /7.3.5 > this is then exported to all of the Unix servers. > > The problem I have with is the Unix Oracle home directory that I also > created in /softwr/oracle/home, ie. > oracle:x:600:211:Oracle user:/softwr/oracle/home:/bin/ksh > but the DBAs are telling me that the oracle home directory > needs to be unique > for each Unix server and Database running on that server because the > home directory is the place where they keep their environment. > > How do other people deal with this? Do you create specific Unix > accounts for each Oracle database running on a server? > > Thanks > George > > -------------------------------------------------------------- > ----------------- > George Kahler e-mail: > george(a)yorku.ca > Sr. Systems Administrator humans: (416) > 736-2100 x.22699 > Computing and Network Services machines: (416) 736-5830 > Ontario, Canada, M3J-1P3 >

1 0

RE: cluster using F740 and F760
by Traitel, Eyal 09 Mar '01

09 Mar '01

No, sorry. This is unsupported. Eyal. ---------------------------------------------------------------------- eTraitel - I'm the new eBuzzword around !!! ---------------------------------------------------------------------- Eyal Traitel - Filer Escalation Engineer CNA, MCSE, CSA, NACA Network Appliance BV Holland Office Center Kruisweg 799b 2132 NG, Hoofddorp The Netherlands Office: +31 23 567 9685 Cellular: +31 6 5497 2568 Email: eyal(a)netapp.com ---------------------------------------------------------------------- Get answers NOW! - NetApp On the Web - http://now.netapp.com ---------------------------------------------------------------------- -----Original Message----- From: max chan [mailto:maxchan0@yahoo.com] Sent: Tuesday, March 06, 2001 7:16 AM To: toasters(a)mathworks.com Subject: cluster using F740 and F760 Dear all, Is it possible to use F740 and F760 as a cluster pair? Max __________________________________________________ Do You Yahoo!? Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/

2 1

Multiple Oracle DBs on a filer
by George Kahler 09 Mar '01

09 Mar '01

Hi, I'm setting up an 840 filer for our DBA group to run about 4-to-5 Oracle OLT databases and 4-to-5 Oracle DSS databases being served from about 1/2 a dozen Unix servers. This is something new to the DBAs in that they are not experienced with NFS or anything like a filer; a common disk storage. Their databases are married to specific Unix servers. The project mandates that the environment needs to be set up in such a way that if any Unix server fails, the database service that it was providing could easily be brought up on any other available Unix server. (So far no problem) I have created and mapped out a quota tree into which I installed the Oracle distribution. It looks something like: /softwr/oracle/product/7.3.4 /7.3.5 this is then exported to all of the Unix servers. The problem I have with is the Unix Oracle home directory that I also created in /softwr/oracle/home, ie. oracle:x:600:211:Oracle user:/softwr/oracle/home:/bin/ksh but the DBAs are telling me that the oracle home directory needs to be unique for each Unix server and Database running on that server because the home directory is the place where they keep their environment. How do other people deal with this? Do you create specific Unix accounts for each Oracle database running on a server? Thanks George ------------------------------------------------------------------------------- George Kahler e-mail: george(a)yorku.ca Sr. Systems Administrator humans: (416) 736-2100 x.22699 Computing and Network Services machines: (416) 736-5830 Ontario, Canada, M3J-1P3

3 2

NDMP restore not reflected in sysstat?
by Tom "Mad Dog" Yergeau 08 Mar '01

08 Mar '01

Is the data written to disk during an NDMP restore not reflected in sysstat? If not, why not? (See sysstat output below. About 8MB/sec is coming in from tape, but the write to disk is never reflected in sysstat) Thanks, Tom CPU NFS CIFS HTTP Net kB/s Disk kB/s Tape kB/s Cache in out read write read write age 29% 0 878 0 597 583 522 0 8809 0 27 38% 0 1535 0 294 503 348 0 8589 0 27 33% 0 1503 0 380 252 176 0 8516 0 27 21% 0 772 0 173 236 212 0 8847 0 27 26% 0 691 0 132 251 360 0 8903 0 27 26% 0 921 0 190 351 192 0 8709 0 27 22% 0 790 0 161 226 176 0 8589 0 27 27% 0 1140 0 246 905 344 0 8387 0 27 19% 0 677 0 138 139 268 0 8007 0 27 24% 0 732 0 148 139 236 992 8193 0 27 30% 0 760 0 178 651 996 3715 8847 0 27 20% 0 767 0 150 167 212 0 7870 0 27 20% 0 792 0 147 147 184 0 8266 0 27 20% 0 703 0 135 143 192 0 8709 0 27 23% 0 798 0 161 146 176 0 8524 0 27 22% 0 845 0 171 177 168 0 8580 0 27 22% 0 834 0 157 172 232 0 8847 0 27 33% 0 1350 0 246 231 188 0 8580 0 27 36% 0 1515 0 283 351 220 0 8589 0 27 CPU NFS CIFS HTTP Net kB/s Disk kB/s Tape kB/s Cache in out read write read write age 32% 0 974 0 182 216 480 0 8516 0 27 40% 0 1254 0 217 308 358 0 8297 0 27 40% 0 1416 0 261 257 566 2112 8809 0 27 34% 0 1432 0 253 250 188 0 8524 0 27 38% 0 1450 0 264 298 212 0 8709 0 27 36% 0 1418 0 263 265 184 0 8912 0 27 37% 0 1581 0 288 273 156 0 8774 0 27 38% 0 1610 0 303 295 160 0 8709 0 27 35% 0 1459 0 273 261 188 0 8460 0 27 37% 0 1622 0 306 274 124 0 8645 0 27 41% 0 1756 0 341 341 180 0 8782 0 27 40% 0 1644 0 308 308 200 0 8451 0 27 40% 0 1473 0 284 258 123 0 7783 0 27 34% 0 1255 0 238 220 357 1600 8429 0 27 26% 0 974 0 257 282 136 0 8718 0 27 49% 0 1459 0 291 950 1336 0 8645 0 27 38% 0 1268 0 259 977 680 0 8718 0 27 39% 0 1411 0 262 238 352 0 8709 0 27 40% 0 1293 0 280 266 248 0 8589 0 27 42% 0 1676 0 360 1277 204 0 8838 0 27

4 3

RSH problem on NT 4.0
by Uros Lampret 08 Mar '01

08 Mar '01

Hi! I have problems performing RSH from specific servers. This is from NT 4.0 SP6. The first RSH goes OK, the second one hangs. Every time! If a wait some time between RSH command it works OK. This is Exchange server. I tried it from five diferent servers and workstations and they all work fine. I'd like to run batch to delete old snapshot, rename and create new. It always execute only first command and the others time out. Only on this server? Any ideas? Thanks! Microsoft(R) Windows NT(TM) (C) Copyright 1985-1996 Microsoft Corp. C:\>rsh ntnas1 df Filesystem kbytes used avail capacity Mounted on /vol/vol0/ 101684188 3429404 98254784 3% /vol/vol0/ /vol/vol0/.snapshot 25421044 390456 25030588 2% /vol/vol0/.snapsh ot C:\>rsh ntnas1 version -> ntnas1:Connection timed out rsh: can't establish connection _____________________________ Uros Lampret www.ourspace.si _____________________________

4 3

ndmp ops timestamps are wrong.
by neil lehrer 08 Mar '01

08 Mar '01

hi, we are running tests for legato and i noticed that the timestamps for ndmp messages on the filer console and in the ndmp log are exactly 7 hours earlier than they should be. the timestamps on the filer console for other ops are correct. tz on filer is est5edt. any ideas. thanks. --

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters