11 messages in com.mysql.lists.clusterRe: Performance with cluster under 5....| From | Sent On | Attachments |
|---|---|---|
| Dan Trainor | 24 Feb 2006 14:15 | |
| Dan Trainor | 24 Feb 2006 14:16 | |
| Kishore Jalleda | 24 Feb 2006 14:26 | |
| Dan Trainor | 24 Feb 2006 14:28 | |
| Kishore Jalleda | 24 Feb 2006 14:48 | |
| Dan Trainor | 24 Feb 2006 15:01 | |
| Stewart Smith | 26 Feb 2006 16:00 | |
| Stewart Smith | 27 Feb 2006 03:45 | |
| Cor...@ SkyVantage | 03 Mar 2006 09:35 | |
| Jimmy Guerrero | 03 Mar 2006 09:39 | |
| Cor...@ SkyVantage | 03 Mar 2006 09:44 |
| Subject: | Re: Performance with cluster under 5.0.18![]() |
|---|---|
| From: | Dan Trainor (da...@id-confirm.com) |
| Date: | 02/24/2006 03:01:15 PM |
| List: | com.mysql.lists.cluster |
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Kishore Jalleda wrote:
Hi Dan, Looks like you are not very inclined towards replication, but have you had a chance to test Multi-Master or a Multi-Master Replication Ring and checked its performance, I have a Dual-Master setup, and the replication is pretty fast ensuring good data consistency, but as you said it's all in the air now and who knows!!!!!
Have a good weekend
Kishore Jalleda
Thanks for the response, Kishore -
Yes, and I agree with you on that point. I understand that Cluster's goal is not speed. Although I'm sure they're working on it, I don't expect Cluster to ever match the speed of a standalone - or even replicating - system.
The reason for using Cluster in this implementation is to ensure data consistency throughout the database, where I can distribute reads and writes. My testing, and studies, have shown that although replication is quite a bit faster, it does not provide the data consistancy guarantee that I require - even with a multi-master setup.
Who knows - it's still up in the air now ;)
Thanks! - -dant -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFD/4jBdJpmX+LLzdoRAlWoAKCIhhiOLxjUu4qWmT0EviD1A3+5HwCfa5vf SyfYCPSdGfh/2cZaTDMHpVY= =8oyK -----END PGP SIGNATURE-----
On 2/24/06, Dan Trainor <da...@id-confirm.com> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Kishore Jalleda wrote:
Yes I do agree that some tuning might increase the performance of the cluster a bit, but my guess is not by a whole lot, and also as I said in an earlier thread that a cluster would not necessarily give you speed ( a single mysql instance would definitely outperform a cluster for speed ---obviously coz of lesser overhead) but a cluster would give you scalability, basically it increases your concurrency at which you can service clients, so a cluster may not be a good option in your case, I would suggest that you use replication instead and have multiple slaves and distribute the load, if you have the resources to do that ......
Kishore Jalleda
On 2/24/06, Dan Trainor <da...@id-confirm.com> wrote: Oh yeah -
It looks like this thread was one of the more helpful in explaining a few things:
http://forums.mysql.com/read.php?25,34915,34915#msg-34915
Thanks to Josh for his feedback, that helped a lot. Unfortunately the named resolution for that problem, is not helping me.
Thanks! -dant
- -- MySQL Cluster Mailing List For list archives: http://lists.mysql.com/cluster To unsubscribe: http://lists.mysql.com/cluster?unsub=kjal...@gmail.com
Hello, Kishore -
Yes, I have done extensive testing with multi-master, and it is impressive. However, I don't believe that it carries the scalability that Cluster has, which is why I was interested in cluster.
You have a good weekend, too. Thanks.
- -dant -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFD/5BSdJpmX+LLzdoRAjUaAJ40HGx9uS/DVDlt333mwKLsSlliQQCghsJv Ie69UUxFxipaGIeZqHWd9OA= =ei6n -----END PGP SIGNATURE-----




