Differences between version 15 and predecessor to the previous major change of PgBench.
Other diffs: Previous Revision, Previous Author, or view the Annotated Edit History
Newer page: | version 15 | Last edited on Thursday, June 15, 2006 8:05:53 pm | by AristotlePagaltzis | Revert |
Older page: | version 3 | Last edited on Thursday, June 15, 2006 4:11:18 pm | by PerryLorier | Revert |
@@ -1,5 +1,5 @@
-This is a scratch pad for some [PostgreSQL | http://www.postgresql.org
] benchmarks. The contributed utility <tt>pgbench</tt> is used for the testing.
+This is a scratch pad for some [PostgreSQL] benchmarks. The contributed utility <tt>pgbench</tt> is used for the testing.
For most of the testing, important parts of the postgres configuration used are:
<verbatim>
shared_buffers = 23987
@@ -12,8 +12,12 @@
checkpoint_warning = 300
commit_delay = 20000
commit_siblings = 3
wal_sync_method = fdatasync
+
+ enable_seqscan = off
+ default_with_oids = off
+ stats_start_collector = false
</verbatim>
Exceptions will be listed as the tests are performed.
@@ -24,13 +28,15 @@
The base hardware is:
HP DL380 G4%%%
Dual 3.20GHz Xeon, 1MB L2 cache, 800MHz FSB, Hyperthreading disabled%%%
- 1GB DDR
-333
(PC2700
) registered ECC memory%%%
+ 1GB DDR2
-400
(PC2-3200
) registered ECC memory%%%
Broadcom PCI-X onboard network adapters%%%
- Battery-backed write cache enabled.
+ ~SmartArray 6i onboard%%%
+
Battery-backed write cache enabled.%%%
- Linux 2.4.27 (from Debian <tt>kernel-image-2.4.27-2-686-smp</tt>)
+ Linux 2.4.27 (from Debian <tt>kernel-image-2.4.27-2-686-smp</tt>)%%%
+ Using ext3 with 'ordered' data mode%%%
On with the testing!
----
@@ -120,8 +126,53 @@
number of transactions per client: 2000
number of transactions actually processed: 200000/200000
tps = 409.561669 (including connections establishing)
tps = 414.078634 (excluding connections establishing)
+ </pre>
+
+----
+
+Data array: RAID1+0, 4x 72GB 15k RPM%%%
+WAL array: On data array%%%
+Other notes: %%%
+
+ <pre>
+ scaling factor: 100
+ number of clients: 100
+ number of transactions per client: 1000
+ number of transactions actually processed: 100000/100000
+ tps = 325.140579 (including connections establishing)
+ tps = 330.843403 (excluding connections establishing)
+ </pre>
+
+----
+
+Data array: RAID5, 4x 72GB 15k RPM%%%
+WAL array: RAID1, 2x 72GB 10k RPM%%%
+Other notes: %%%
+
+ <pre>
+ scaling factor: 100
+ number of clients: 100
+ number of transactions per client: 1000
+ number of transactions actually processed: 100000/100000
+ tps = 236.721312 (including connections establishing)
+ tps = 239.738377 (excluding connections establishing)
+ </pre>
+
+----
+
+Data array: RAID5, 4x 72GB 15k RPM%%%
+WAL array: On data array%%%
+Other notes: %%%
+
+ <pre>
+ scaling factor: 100
+ number of clients: 100
+ number of transactions per client: 1000
+ number of transactions actually processed: 100000/100000
+ tps = 192.430583 (including connections establishing)
+ tps = 194.404205 (excluding connections establishing)
</pre>
----
@@ -132,5 +183,14 @@
<pre>
</pre>
----
+Other observations
+* The test database started at 1.4GB, and got to at least 14GB during testing, has this growth affected results?
+* The WAL consumes large amounts of kernel page cache. When moving the WAL between devices, when the old files are unlinked,
+ 1/2 of the page cache is freed. The WAL is never read, and written only once, this is as waste!
+* The BBWC makes write performance very erratic
+* The HP SmartArray hardware (or perhaps driver) tends to block reads while there are cached writes occuring. Large read latencies (seconds) results. I have not yet found a way to tune this.
+
+----
+
CategoryDiskNotes