I am about to lease another server to be a backup web/mail server with a replicated mysql database. The current live server is a dual Xeon 2.8Ghz. Would I do better to get another Xeon server or an Opteron based server? My concerns are compatibility of the database and the SSL certificate currently running on the Xeon machine. Are these concerns unfounded?
TIA
Mike
Both processors are x86 based so the database shouldn't be an issue, if at all.
SSL certificates are not tied to a machine at all, they are tied to the common name that they are called as. You can export the SSL keys do different formats with OpenSSL if you require it. Just make sure you have both your public and private keys and any pass phrases you need.
-Bill
----- Original Message ----- From: centos-bounces@centos.org on behalf of Mike Kercher Sent: Thu, 10/26/2006 2:41pm To: CentOS mailing list Subject: [CentOS] Dual Opteron or Dual Xeon?
I am about to lease another server to be a backup web/mail server with a replicated mysql database. The current live server is a dual Xeon 2.8Ghz. Would I do better to get another Xeon server or an Opteron based server? My concerns are compatibility of the database and the SSL certificate currently running on the Xeon machine. Are these concerns unfounded?
TIA
Mike _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Quoting Mike Kercher mike@vesol.com:
I am about to lease another server to be a backup web/mail server with a replicated mysql database. The current live server is a dual Xeon 2.8Ghz. Would I do better to get another Xeon server or an Opteron based server? My concerns are compatibility of the database and the SSL certificate currently running on the Xeon machine. Are these concerns unfounded?
If compatibility of MySQL database and certificates are your only concerns, it doesn't matter what you get. You can even get something with UltraSPARC or Alpha processor. Or you could get Cray supercomputer. It'll just work.
The certificates are normally just text files. You can move them between machines or operating systems, as long as name of machine (and/or IP addresses) match the names and/or IP addresses for which the certificate is issued. This would be whatever is stored in CN and AltSubjectName fields in the certificate. You can use "openssl x509 -in certfile.txt -noout -text" to dump out certificate in human readable format.
MySQL is just an application. If the hardware and OS are supported by MySQL, you are fine. Linux is supported on all Intel and AMD processors. So you are fine there.
There is only one thing to watch out for both Xeon and Opteron. If your current Xeon is 32-bit, and the new processor (even if it is Xeon) is 64-bit, *and* you install 64-bit version of Linux *and* 64-bit MySQL binaries, you can't just copy binary database files to the new machine. It probably won't work. Well, maybe it would, but I would very strongly recommend against it. Use mysqldump to dump database into file, and re-imort it on new machine. Or setup replication between two machines. This is recommended even when moving database between two identical machines anyhow.
There is only one thing to watch out for both Xeon and Opteron. If your current Xeon is 32-bit, and the new processor (even if it is Xeon) is 64-bit, *and* you install 64-bit version of Linux *and* 64-bit MySQL binaries, you can't just copy binary database files to the new machine. It probably won't work. Well, maybe it would, but I would very strongly recommend against it. Use mysqldump to dump database into file, and re-imort it on new machine. Or setup replication between two machines. This is recommended even when moving database between two identical machines anyhow.
???
I have copied a binary mysql database from an Intel 32-bit box (Pentium 4 2.8Ghz Centos 4.4) to an AMD 64-bit box (AMD Athlon 64 3000+ Centos 4.4 64-bit) without any problems. It was a straight copy of /var/lib/mysql to the other box and then start the mysql server. I did not even copy the my.cnf file since I had not make any changes to the one provided by the rpm. I believe you mean postgresql databases. I had to do psqldump to transfer the postgresql database to the AMD64 box. I think you can even move mysql databases from Windows to Linux without problems.
Aren't mysql databases machine independent?
Mike Kercher wrote:
I am about to lease another server to be a backup web/mail server with a replicated mysql database. The current live server is a dual Xeon 2.8Ghz. Would I do better to get another Xeon server or an Opteron based server? My concerns are compatibility of the database and the SSL certificate currently running on the Xeon machine. Are these concerns unfounded?
You will do better to get an Opteron based server. Opterons handle SSL faster, in fact much faster, than Intel Xeon processors not to mention that I/O on an Opteron box is better too. Stay away from Woodcrest Xeon servers. The US Government and a large korean ISP have had stability problems with Woodcrest servers.
On Fri, 27 Oct 2006 at 12:15pm, Feizhou wrote
You will do better to get an Opteron based server. Opterons handle SSL faster, in fact much faster, than Intel Xeon processors not to mention that I/O on an Opteron box is better too. Stay away from Woodcrest Xeon servers. The US Government and a large korean ISP have had stability problems with Woodcrest servers.
Do you have any more details on said stability problems (e.g. hardware specs, etc)? I've been very impressed with the Woodcrest Xeon tester I've got -- in fact, it bests my Opterons in almost every benchmark I throw at it, and it's been rock solid. I'm leaning pretty hard in that direction for my next cluster upgrade.
Do you have any more details on said stability problems (e.g. hardware specs, etc)? I've been very impressed with the Woodcrest Xeon tester I've got -- in fact, it bests my Opterons in almost every benchmark I throw at it, and it's been rock solid. I'm leaning pretty hard in that direction for my next cluster upgrade.
The information about the US Government throwing the Woodcrest solution out was 'reported' by the Inquirer. :D Perhaps IBM boxes...they also posted another interesting bit about Woodcrest and Intel 965 motherboards having problems with raid5...kind of like how there is bad mixing if you use ext3 + 3ware in RAID5 mode maybe. I am afraid that is all I can give you on this.
The Korean portal one can be found below but it is entirely in korean.
http://www.inews24.com/php/news_view.php?g_menu=020200&g_serial=225735
Something about overheating and FB-DIMM problems which was only encountered after the thing went live.
How close to live are your benchmarks?
On Mon, 30 Oct 2006 at 4:24pm, Feizhou wrote
Do you have any more details on said stability problems (e.g. hardware specs, etc)? I've been very impressed with the Woodcrest Xeon tester I've got -- in fact, it bests my Opterons in almost every benchmark I throw at it, and it's been rock solid. I'm leaning pretty hard in that direction for my next cluster upgrade.
The information about the US Government throwing the Woodcrest solution out was 'reported' by the Inquirer. :D Perhaps IBM boxes...they also posted another interesting bit about Woodcrest and Intel 965 motherboards having problems with raid5...kind of like how there is bad mixing if you use ext3 + 3ware in RAID5 mode maybe. I am afraid that is all I can give you on this.
Ah, yes, the Inquirer. Is there enough salt in the world?
The Korean portal one can be found below but it is entirely in korean.
http://www.inews24.com/php/news_view.php?g_menu=020200&g_serial=225735
Something about overheating and FB-DIMM problems which was only encountered after the thing went live.
Hrm. Sounds like an implementation problem.
How close to live are your benchmarks?
This is what I've got so far http://www.duke.edu/~jlb17/optxeon.pdf. Yes, the Xeons have a slight clockspeed advantage, but not near enough to account for the performance increase. The thermal simulation results surprised me -- I really expected the Opterons to fare better there, given how memory intensive they are. Only the heart phantom sim showed better performance on Opteron due to the onboard memory controller.
Ah, yes, the Inquirer. Is there enough salt in the world?
:D
The Korean portal one can be found below but it is entirely in korean.
http://www.inews24.com/php/news_view.php?g_menu=020200&g_serial=225735
Something about overheating and FB-DIMM problems which was only encountered after the thing went live.
Hrm. Sounds like an implementation problem.
Well...what other chipset provider is there for the Woodcrest? Unless you want to point the extra power needed for FB-DIMM solutions and thus extra cooling I suppose...
How close to live are your benchmarks?
This is what I've got so far http://www.duke.edu/~jlb17/optxeon.pdf. Yes, the Xeons have a slight clockspeed advantage, but not near enough to account for the performance increase. The thermal simulation results surprised me -- I really expected the Opterons to fare better there, given how memory intensive they are. Only the heart phantom sim showed better performance on Opteron due to the onboard memory controller.
The Korean portal also got very good performance figures (better than their Opteron boxes) from the Woodcrest boxes. It was big news for IT in Korea because the portal was exclusively using AMD Opteron.
If you do not get any problems at all with your Supermicro board and Woodcrest then by all means :)
On Mon, 30 Oct 2006 at 10:16pm, Feizhou wrote
Something about overheating and FB-DIMM problems which was only encountered after the thing went live.
Hrm. Sounds like an implementation problem.
Well...what other chipset provider is there for the Woodcrest? Unless you want to point the extra power needed for FB-DIMM solutions and thus extra cooling I suppose...
That (power and cooling) is indeed what I was referring to.
Well...what other chipset provider is there for the Woodcrest? Unless you want to point the extra power needed for FB-DIMM solutions and thus extra cooling I suppose...
That (power and cooling) is indeed what I was referring to.
The old not enough juice causing instability problem eh? So much for Intel's performance per watt dribble. I heard this Dell guy promote Dell's Intel boxes on the basis of performance per watt. I cannot wait to see him do it for Opteron boxes :D.
-
Aleksandar Milivojevic -
Bill Church -
Feizhou -
Joshua Baker-LePain -
Mike Kercher