[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[no subject]
- <!--x-content-type: text/plain -->
- <!--x-date: Sat Feb 14 22:28:23 2004 -->
- <!--x-from-r13: gubznfjbbq ng znp.pbz (Fubznf Ibbq) -->
- <!--x-message-id: E0BC6273-5F66-11D8-9E7B-[email protected] -->
- <!--x-reference: [email protected] -->
- <!--x-reference: A0455D86-5F64-11D8-9612-[email protected] --> "http://www.w3.org/TR/html4/loose.dtd">
- <!--x-subject: [ale] Any MySQL Experts around? -->
- <li><em>date</em>: Sat Feb 14 22:28:23 2004</li>
- <li><em>from</em>: thomaswood at mac.com (Thomas Wood)</li>
- <li><em>in-reply-to</em>: <<a href="msg00503.html">[email protected]</a>></li>
- <li><em>references</em>: <<a href="msg00500.html">[email protected]</a>> <<a href="msg00503.html">[email protected]</a>></li>
- <li><em>subject</em>: [ale] Any MySQL Experts around?</li>
hope that provides some comparison.
wood
On Feb 14, 2004, at 10:11 PM, Jason Etheridge wrote:
>> I'm thinking of having all my server logs (including NT/2K/2K+3)
>> record
>> into a MySQL db for better reporting. Does anyone know if MySQL can
>> handle 100,000+ inserts per day? Also, if it handling this many
>> inserts,
>> how good is the SELECT performance for the automated reporting? ANy
>> thoughts, pointers?
>
> Disclaimer: I'm not a real DBA and the tuple calculus I learned in
> school actually tripped me up in the real world :D
>
> The largest table I work with in mysql has roughly 8 million records
> using full-text indexing. Each row contains a little over 260 bytes
> of data. Every morning I turn indexing off and do a LOAD DATA REPLACE
> touching every record, changing a good bit of them, and adding a
> thousand or so new records. It can take around 3 hours to do that and
> reindex on an old Sun E10K (I forget how many processors; maybe a
> dozen for that domain). Queries are fast after that, but I use READ
> LOCK's and there are no updates during the day. I'm not reporting on
> live data.
>
> If you don't use full-text indexing, and instead use LIKE statements
> on large fields, searching is going to be pretty slow with that many
> records, especially if you can't lock the tables because you're doing
> updates. If you do use full-text indexing, I'd worry about 100,000+
> inserts a day and concurrent searching. Full-text updates work best
> with indexing disabled, and even then, when you go to renable
> indexing, it'll do it via REPAIR BY SORTING, which could take a long
> time if you don't have enough RAM to do it without thrashing the
> disks.
>
> What sort of reporting do you want to do and sort of logs are they?
> If you just want to store key statistical information rather than
> entire log entries, that would be better. But for storing actual
> logs, I'm not sure relational databases are the best tool. For
> log-based reporting, I currently use a kludge: the system loads into
> temporary tables specific data parsed from log entries from just the
> months we're interested in for the current running report. This only
> works well for one or two months in our case because of the arbitrary
> timeout value I set. For anything really large, like a year's worth
> of transactions, I just run the logs through a perl script and forget
> mysql.
>
> I have another table without full-text indexing that gets searched and
> updated throughout the day, with about a million records. But nothing
> is added except in the mornings, just updated. I'll try to find some
> stats for that if you want.
>
> Best thing to do is get a test machine and play around with some
> sample loads, and make sure to allocate a lot of filehandles for
> mysql's process space.
>
> -- Jason
>
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> <a rel="nofollow" href="http://www.ale.org/mailman/listinfo/ale">http://www.ale.org/mailman/listinfo/ale</a>
>
</pre>
<!--X-Body-of-Message-End-->
<!--X-MsgBody-End-->
<!--X-Follow-Ups-->
<hr>
<!--X-Follow-Ups-End-->
<!--X-References-->
<ul><li><strong>References</strong>:
<ul>
<li><strong><a name="00500" href="msg00500.html">[ale] Any MySQL Experts around?</a></strong>
<ul><li><em>From:</em> jonathan.glass at ibb.gatech.edu (Jonathan Glass)</li></ul></li>
<li><strong><a name="00503" href="msg00503.html">[ale] Any MySQL Experts around?</a></strong>
<ul><li><em>From:</em> phasefx at magusaptus.com (Jason Etheridge)</li></ul></li>
</ul></li></ul>
<!--X-References-End-->
<!--X-BotPNI-->
<ul>
<li>Prev by Date:
<strong><a href="msg00503.html">[ale] Any MySQL Experts around?</a></strong>
</li>
<li>Next by Date:
<strong><a href="msg00509.html">[ale] Any MySQL Experts around?</a></strong>
</li>
<li>Previous by thread:
<strong><a href="msg00503.html">[ale] Any MySQL Experts around?</a></strong>
</li>
<li>Next by thread:
<strong><a href="msg00509.html">[ale] Any MySQL Experts around?</a></strong>
</li>
<li>Index(es):
<ul>
<li><a href="maillist.html#00504"><strong>Date</strong></a></li>
<li><a href="threads.html#00504"><strong>Thread</strong></a></li>
</ul>
</li>
</ul>
<!--X-BotPNI-End-->
<!--X-User-Footer-->
<!--X-User-Footer-End-->
</body>
</html>