2019. 01. 24. 9:50 keltezéssel, Gabor Boros írta:
I will execute my ISQL cases with 2.5 SS and 2.5 SC on a real HDD. We
will see...
- 1 ISQL 1 DB
2.5 SS - 3:03
2.5 SC - 3:08
- 5 ISQL 1 DB
2.5 SS - 11:13
2.5 SC - 7:28
- 5 ISQL 5 DB
2.5 SS - 11:11
2.5 SC - 7:22
Gabor
Firebird-Devel
On Thu, 24 Jan 2019 10:49:14 +0300
Roman Simakov wrote:
> чт, 24 янв. 2019 г. в 10:25, liviuslivius
> :
> >
> > Hi.
> >
> > I have thinked about oltp comparision, and i think that comparing
> > e.g. classic vs superserver on RAM disc is wrong. Why? Because read
> > from "disc" is as fast as read
On 24/01/2019 09:01, Doychin Bondzhev wrote:
Many small installations are on hdd.
My first reaction was that HDD is still the more common base for
systems, and for large volumes of data I think that will be the case for
a while yet? But I've been adding SSD's as the boot disk on my servers
V3 super server should be compared with V2.5 super classic.
Slavek
2.5 - 10:55
3.0 - 8:11
4.0 - 8:41
Same script with numbered(1..5) table names into one database
concurrently:
2.5 - 10:51 - 336.72 MB
3.0 - 8:09 - 329.79 MB
4.0 - 8:45 - 329.79 MB
i.e. v3 finally becomes faster than
2019. 01. 24. 8:27 keltezéssel, Dmitry Yemanov írta:
i.e. v3 finally becomes faster than v2.5, but v4 is slower than v3
(while faster than v2.5)
Correct?
Yes. But 2.5 SC beats them all. I made a graph from the results. The
numbers are seconds and the smaller is the better.
Gabor
Many small installations are on hdd.
On Thu, Jan 24, 2019, 10:57 Alex Peshkoff via Firebird-devel <
firebird-devel@lists.sourceforge.net wrote:
> On 1/24/19 11:50 AM, Gabor Boros wrote:
> > 2019. 01. 24. 8:24 keltezéssel, liviuslivius írta:
> >> Hi.
> >>
> >> I have thinked about oltp
On 1/24/19 11:50 AM, Gabor Boros wrote:
2019. 01. 24. 8:24 keltezéssel, liviuslivius írta:
Hi.
I have thinked about oltp comparision, and i think that comparing
e.g. classic vs superserver on RAM disc is wrong. Why? Because read
from "disc" is as fast as read from cache.
The comparision can
On 1/24/19 10:24 AM, liviuslivius wrote:
Because read from "disc" is as fast as read from cache.
That's worth testing for particular SSD. For example my SSD
(WDS240G1G0B) works a few times slower than RAM disk.
Firebird-Devel mailing list, web interface at
2019. 01. 24. 8:24 keltezéssel, liviuslivius írta:
Hi.
I have thinked about oltp comparision, and i think that comparing e.g.
classic vs superserver on RAM disc is wrong. Why? Because read from
"disc" is as fast as read from cache.
The comparision can show some problems but results must be
чт, 24 янв. 2019 г. в 10:25, liviuslivius :
>
> Hi.
>
> I have thinked about oltp comparision, and i think that comparing e.g.
> classic vs superserver on RAM disc is wrong. Why? Because read from "disc" is
> as fast as read from cache.
This case eliminates disc problems and allow to highlight
23.01.2019 17:50, Gabor Boros wrote:
Same script into 5 databases concurrently:
2.5 - 10:55
3.0 - 8:11
4.0 - 8:41
Same script with numbered(1..5) table names into one database concurrently:
2.5 - 10:51 - 336.72 MB
3.0 - 8:09 - 329.79 MB
4.0 - 8:45 - 329.79 MB
i.e. v3 finally becomes
Hi.
I have thinked about oltp comparision, and i think that comparing e.g. classic
vs superserver on RAM disc is wrong. Why? Because read from "disc" is as fast
as read from cache. The comparision can show some problems but results must be
analysed carefully.
Regards,Karol Bieniaszewski
I got a request privately about the 2.5 SuperClassic numbers. The
numbers are (with DefaultDbCachePages = 1):
The original case:
3:03 - 65.56 MB
Same script into 5 databases concurrently:
7:21
Same script with numbered(1..5) table names into one database concurrently:
7:22 -
2019. 01. 21. 17:41 keltezéssel, Dmitry Yemanov írta:
This is a single-threaded test, isn't it?
2.5 - 2:59 - 65.65 MB
3.0 - 3:27 - 68.95 MB
4.0 - 3:37 - 68.95 MB
The results are more or less expected in this case (it doesn't mean
they're desirable but this is a different story).
Same
19.01.2019 23:14, Gabor Boros wrote:
The ISQL case is. I made a new database (CREATE DATABASE
'R:\DB\INSERT_TEST.FDB' PAGE SIZE 8192;) with 2.5, 3.0 and 4.0. Then
execute a one table script on it with ISQL (isql.exe -u SYSDBA -p
masterkey 192.168.0.99:R:\DB\INSERT_TEST.FDB -i DATA.sql).
2019. 01. 21. 10:51 keltezéssel, Pavel Cisar írta:
Hi,
I think that there is enough evidence (provided from various sources,
incl. this comparison) that 2.5 is still "significantly" (i.e. not in
range that should be ignored) faster than 3.0. If 4.0 is even worse than
3.0, it's another
Hi All,
I wanted to know how much faster 3.0 is than 2.5 and 4.0 than 3.0. Tried
to compare performance of 2.5, 3.0 and 4.0 with Firebird's OLTP solution
(executed ~100 times). The results are constant for me. 3.0 is slower
than 2.5 and 4.0 is slower than 3.0. (Some numbers published in the
17 matches
Mail list logo
