1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-25 23:21:54 +03:00
Commit Graph

34 Commits

Author SHA1 Message Date
Andrew Tridgell
dc4f1f6754 r23535: accept numeric ntstatus codes in nbench load files (ronnies wireshark
script generates numeric codes)
(This used to be commit 455895c4b1)
2007-10-10 14:53:25 -05:00
Andrew Tridgell
551575012b r23534: honour the targetrate even for loadfiles with timestamps
(This used to be commit 24b6aabd11)
2007-10-10 14:53:25 -05:00
Andrew Tridgell
2a3282f872 r23533: added --option torture:targetrate=rate to target a particular
throughput in MBytes/sec
(This used to be commit 35477e8e3d)
2007-10-10 14:53:25 -05:00
Andrew Tridgell
428373743a r23531: added optional latency reporting in BENCH-NBENCH. To use it, you need
to have a load file that puts a timestamp in the first column. That
tells NBENCH to target the same throughput as the capture, and to
report the maximum amount of latency miss in the test

This allows you to quickly see how many clients you can run with while
still meeting a given target
(This used to be commit 7f045ca4f5)
2007-10-10 14:53:24 -05:00
Andrew Tridgell
1c211a2e43 r20650: revert a bunch of code I didn't mean to commit yet
(This used to be commit b3e2d49087)
2007-10-10 14:37:26 -05:00
Andrew Tridgell
f6274959ba r20647: add cluster code
(This used to be commit 5870830b99)
2007-10-10 14:37:24 -05:00
Andrew Tridgell
c4274fee71 r20302: added support for 64 bit file offsets in NBENCH
(This used to be commit 9a9f189141)
2007-10-10 14:29:35 -05:00
Andrew Tridgell
114265b4d5 r20301: add an option torture:readonly which allows the NBENCH test to be run
read only. This is only any use with customer loadfiles, and
pre-created data. I'm using this to test performance as a video
server, with pre-created video files.
(This used to be commit 48a322a9c1)
2007-10-10 14:29:35 -05:00
Jelmer Vernooij
108de297a7 r19510: Make smbtorture's `nprocs' parameter a parametric option.
(This used to be commit f5d6b3ca29)
2007-10-10 14:24:42 -05:00
Jelmer Vernooij
a39f239cb2 r19392: Use torture_setting_* rather than lp_parm_* where possible.
(This used to be commit b28860978f)
2007-10-10 14:21:24 -05:00
Jelmer Vernooij
8773e743c5 r19339: Merge my 4.0-unittest branch. This adds an API for more fine-grained
output in the testsuite rather than just True or False for a
set of tests.

The aim is to use this for:
 * known failure lists (run all tests and detect tests that
						started working or started failing). This
   would allow us to get rid of the RPC-SAMBA3-* tests
 * nicer torture output
 * simplification of the testsuite system
 * compatibility with other unit testing systems
 * easier usage of smbtorture (being able to run one test
						and automatically set up the environment for that)

This is still a work-in-progress; expect more updates over the next couple of
days.
(This used to be commit 0eb6097305)
2007-10-10 14:21:12 -05:00
Volker Lendecke
b7c5bc522b r16907: Add an index parameter to torture_open_connection. Next step is to enable the
unclist parameter for all tests that do two connections, to enable cluster
testing.

Volker
(This used to be commit a5d6db0924)
2007-10-10 14:09:58 -05:00
Jelmer Vernooij
845df42a19 r16333: Move more code out of the core smbtorture. It now no longer
contains protocol-specific code.
(This used to be commit 819d3b4576)
2007-10-10 14:09:11 -05:00
Jelmer Vernooij
909b111f58 r14720: Add torture_context argument to all torture tests
(This used to be commit 3c7a5ce291)
2007-10-10 13:59:13 -05:00
Jelmer Vernooij
d09b70c98b r14527: Fix build problems.
(This used to be commit 863ca4014d)
2007-10-10 13:57:37 -05:00
Stefan Metzmacher
1a53c1dc92 r13346: use private proto header files for the torture tests
metze
(This used to be commit 67837dbd2b)
2007-10-10 13:51:47 -05:00
Jelmer Vernooij
78c50015bb r12694: Move some headers to the directory of the subsystem they belong to.
(This used to be commit c722f665c9)
2007-10-10 13:49:39 -05:00
Jelmer Vernooij
25bb00fbcd r12693: Move core data structures out of smb.h into core.h
torture prototypes in seperate header
(This used to be commit 73610639b2)
2007-10-10 13:49:39 -05:00
Jelmer Vernooij
e3b8d856aa r10662: Eliminate pstring and friends.
(This used to be commit ecef1ee6cd)
2007-10-10 13:39:15 -05:00
Andrew Tridgell
34f48e0e09 r7431: this should fix the bug that mkaplan and I noticed which is that
BENCH-NBENCH sometimes incorrectly reports that it failed.
(This used to be commit 167c29bec6)
2007-10-10 13:17:52 -05:00
Jeremy Allison
2825b0cddf r7180: Missed fixing the shutdown for 1 process...
Jeremy.
(This used to be commit 997e1405cb)
2007-10-10 13:17:23 -05:00
Jeremy Allison
cbf0a4c362 r7178: If we're using one process in NETBENCH simultion only open one
connection to the server. Makes valgrind/cachegrind on the server
easier.
Jeremy.
(This used to be commit fba9462591)
2007-10-10 13:17:23 -05:00
Andrew Tridgell
e82aad1ce3 r5298: - got rid of pstring.h from includes.h. This at least makes it a bit
less likely that anyone will use pstring for new code

 - got rid of winbind_client.h from includes.h. This one triggered a
   huge change, as winbind_client.h was including system/filesys.h and
   defining the old uint32 and uint16 types, as well as its own
   pstring and fstring.
(This used to be commit 9db6c79e90)
2007-10-10 13:09:38 -05:00
Andrew Tridgell
b6b0106427 r3976: changed NBENCH to use the same recording method as the latest dbench,
where the warmup phase continues until all clients have done some file
IO. This gives more repeatable results when under high load
(This used to be commit aca0658f6d)
2007-10-10 13:06:10 -05:00
Andrew Tridgell
dde0705807 r3507: - added deferred replies on sharing violation in pvfs open. The
deferred reply is short-circuited immediately when the file is
  closed by another user, allowing it to be opened by the waiting user.

- added a sane set of timeval manipulation routines

- converted all the events code and code that uses it to use struct
  timeval instead of time_t, which allows for microsecond resolution
  instead of 1 second resolution. This was needed for doing the pvfs
  deferred open code, and is why the patch is so big.
(This used to be commit 0d51511d40)
2007-10-10 13:05:23 -05:00
Andrew Tridgell
ba6d5fcb97 r3324: made the smbtorture code completely warning free
(This used to be commit 7067bb9b52)
2007-10-10 13:04:55 -05:00
Jim McDonough
9529b4948e r3040: Add sleeps between operations to nbench. Submitted by aliguori@us.ibm.com.
(This used to be commit b1bbf0a431)
2007-10-10 13:00:01 -05:00
Stefan Metzmacher
c5fbb6f23c r1654: rename cli_ -> smbcli_
rename CLI_ -> SMBCLI_

metze
(This used to be commit 8441750fd9)
2007-10-10 12:57:47 -05:00
Andrew Bartlett
b1268fc445 r1123: Make all lp_ string functions return 'const char *'.
Fix other 'const' warnings in the torture code.

Andrew Bartlett
(This used to be commit 5d39d7497f)
2007-10-10 12:56:38 -05:00
Andrew Tridgell
516e90bea9 use client1, client2 not client0, client1 to better match netbench
(This used to be commit 4bbc77fa0c)
2004-03-26 02:09:51 +00:00
Andrew Tridgell
3d5ea8e3e2 nicer way of handling \clients
(This used to be commit 931138cc44)
2004-03-10 09:00:24 +00:00
Andrew Tridgell
98cbc669a1 added a 5% warmup time to NBENCH to make the results more consistent
(This used to be commit df50911a9c)
2004-03-10 03:24:03 +00:00
Andrew Tridgell
dda01cea55 handle incomplete load files
change default timelimit to 600 secs
(This used to be commit c5c75a6ada)
2004-03-09 02:24:14 +00:00
Andrew Tridgell
2d9a8bf0c9 greatly improved NBENCH netbench simulator
- more accurate request simulation using Samba4 raw client library
 - now checks for correct status on all requests
 - timelimit based benchmarking, default of 600 seconds
(This used to be commit a46c009c04)
2004-03-09 02:13:13 +00:00