From: Ed Smith <esmith@virtualiron.com>
To: xen-devel@lists.xensource.com
Subject: Testing status of fully virtualized guests (Intel VT) on 32bit XEN unstable
Date: Wed, 10 May 2006 15:51:29 -0400 [thread overview]
Message-ID: <44624441.6040609@virtualiron.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 3305 bytes --]
* NEW: My 32bit XEN builds are still not PAE enabled. Chasing done
why this is. For now I stopped running the 32b SMP Guests.
Test Configuration:
Dell PowerEdge 430, Dual Core, 2GB, 3 SATA (Intel VT)
32bit XEN PAE Hypervisor on a RHEL4U2 32bit root (/dev/sda)
dom0_mem=256M (required to boot domUs)
32bit fully virtualized (HVM) guest RHEL4U2 256MB (/dev/sdb)
pae=0, acpi=1, apic=1
kernargs clock=pit
32bit fully virtualized (HVM) guest RHEL3U6 256MB (/dev/sdc)
pae=0, acpi=1, apic=1
kernargs clock=pit
Boot Tests:
Boot a fully virtualized (HVM) guest to the login prompt
Results are marked Pass|Fail where (n) points to a failure description
Regression Tests:
851 tests (850 ltp tests and one 30 minute user load test)
Tests are marked #Pass/#Fail where (n) points to a failure description
XEN 32bit PAE 2 CPU Hypervisor (booted smp):
----------------------------------------------------------------------
| XEN | Guest Kernel (SMP kernels booted with 2 CPU) |
| Changeset|-----------------------------------------------------------|
| | RHEL4 UP | RHEL4 SMP | RHEL3 UP | RHEL3 SMP |
| |--------------|--------------|--------------|--------------|
| | Boot | Test | Boot | Test | Boot | Test | Boot | Test |
|----------|------|-------|------|-------|------|-------|------|-------|
| 9960 | Pass | 850/1 | | | Pass |Running| | |
| | | (2) | | | | (2,4) | | |
|----------|------|-------|------|-------|------|-------|------|-------|
| 9925 | Pass | 850/1 | Fail | | Pass | 850/1 | Fail | |
| | | (2) | (1) | | | (2,4) | (1) | |
|----------|------|-------|------|-------|------|-------|------|-------|
| 9920 | Pass | 850/1 | Fail | | Pass | 850/1 | Fail | |
| | | (2) | (1) | | | (2,4) | (1) | |
|----------|------|-------|------|-------|------|-------|------|-------|
| 9913 | Fail | | Fail | | Fail | | Fail | |
| | (3) | | (3) | | (3) | | (3) | |
|----------|------|-------|------|-------|------|-------|------|-------|
| 9903 | Fail | | Fail | | Fail | | Fail | |
| | (3) | | (3) | | (3) | | (3) | |
----------------------------------------------------------------------
Failures:
1. 32 bit SMP guests hang on boot
"Uncompressing Linux... Ok, booting the kernel."
2. 32bit UP guest fail ltp gettimeofday02
"Time is going backwards"
3. [Fixed in 9920] Build broken:
cc1: warnings being treated as errors
mm.c: In function subarch_init_memory:
mm.c:163: warning: format %ld expects type long int,
but argument 2 has type unsigned int
mm.c:163: warning: format %ld expects type long int,
but argument 3 has type unsigned int
4. The RHEL3 UP tests run OK however they take 12 hours
to complete where on RHEL4 UP the run for ~4 hours.
There seems to be some performance issue running a
2.4 kernel fully virtualized guest. One test in
particular ltp's aso-stress025 takes 4 hour to run
on RHEL3U6-32b while it take 4 minutes on RHEL4U2-32b.
[-- Attachment #2: regression.2 --]
[-- Type: text/plain, Size: 1819 bytes --]
File: regression.1
Time Level Message
05:00:50 INFO Reporting status: 'Test Running' for test: ltp_gettimeofday02
05:00:52 INFO Preparing to run test 'ltp_gettimeofday02' using profile: /qa/conductor/profiles/ltp/syscalls/gettimeofday02.xml
05:00:52 INFO Starting test 'ltp_gettimeofday02' using profile: /qa/conductor/profiles/ltp/syscalls/gettimeofday02.xml
05:00:52 INFO Dispatching operation: RemoteShell
05:00:52 FINE Client sequencer got message requesting the start of a new test: ltp_gettimeofday02
05:00:52 FINER Client sequencer sent message of type: 4 with seq num: 1 of size: 289 bytes
05:00:52 FINER Client sequencer handling new operation from control sequencer
05:00:52 FINE Client sequencer looking for class: com.katana.conductor.operations.RemoteShell
05:00:52 INFO Operation RemoteShell running
05:00:52 FINE Client sequencer was told that an operation is now running
05:00:52 INFO RemoteShell: target node(s) = vs177
05:00:52 INFO ssh: /usr/bin/ssh root@vs177 cd /qa/conductor/tests/ltp/testcases/bin; gettimeofday02
05:00:52 FINE ssh: waiting for command to finish
05:00:53 INFO ssh: gettimeofday02 0 INFO : checking if gettimeofday is monotonous, takes 30s
05:00:53 INFO ssh: gettimeofday02 1 FAIL : Time is going backwards (old 1145696453.61428 vs new 1145696453.60660!
05:00:53 FINE executeShellCmd(ssh): exit value is 1
05:00:53 SEVERE RemoteShell: command failed with error = 1
05:00:53 SEVERE Operation RemoteShell failed
05:00:53 SEVERE Reporting status: 'Test Failed' for test: ltp_gettimeofday02
05:00:53 FINE Client sequencer detected operation completed with status of: Fail
05:00:53 FINER Client sequencer sent message of type: 5 with seq num: 2 of size: 429 bytes
05:00:53 SEVERE Crash Collection disabled for queue : RHEL4U2-32b-XEN
05:00:53 INFO Cleaning up after test
[-- Attachment #3: regression.4 --]
[-- Type: text/plain, Size: 6026 bytes --]
Queue: RHEL3U6-32b-XEN32
User: QA
Hypervisor Build: 20060505_000 Version: main Config: Release Path: /repository/trees/main/20060505_000/xen32/xen-unstable/dist Dist:Xen32
Completion status:
test run completed
Start time: May 4, 2006 3:57:19 PM
End time: May 5, 2006 4:11:28 AM
Elapsed time: 12 hours 14 minutes 9 seconds
Tests/Config changes in queue: 851
Tests processed: 851
Tests passed: 850
Tests failed: 1
Tests aborted: 0
Forced reboots: 0
aio-stress025 performance issue:
RHEL4U2-32b native hardware: 6 seconds
RHEL4U2-32b domU guest: 3 minutes 31 seconds
RHEL3U6-32b domU guest: 173 minutes 1 second
Here are the timed runs:
dom0 run on native hardware:
----------------------------
[root@tst122 ~]# time ./aio-stress -I500 -o3 -O -r512 -t8 /test/aiodio/junkfile /test/aiodio/file2 /test/aiodio/file7 /test/aiodio/file8 /test/aiodio/file3 /test/aiodio/file4 /test/aiodio/file5 /test/aiodio/file6
adding stage random read
starting with random read
file size 1024MB, record size 512KB, depth 64, ios per iteration 8
max io_submit 64, buffer alignment set to 4KB
threads 8 files 8 contexts 1 context offset 2MB verification off
Running multi thread version num_threads:8
random read on /test/aiodio/file3 (182109.19 MB/s) 1024.00 MB in 0.01s
random read on /test/aiodio/file6 (175975.25 MB/s) 1024.00 MB in 0.01s
random read on /test/aiodio/file4 (176887.20 MB/s) 1024.00 MB in 0.01s
random read on /test/aiodio/file2 (40123.82 MB/s) 1024.00 MB in 0.03s
random read on /test/aiodio/file5 (95934.05 MB/s) 1024.00 MB in 0.01s
random read on /test/aiodio/file8 (4322.75 MB/s) 1024.00 MB in 0.24s
random read on /test/aiodio/file7 (4224.42 MB/s) 1024.00 MB in 0.24s
thread 5 random read totals (2815.35 MB/s) 1024.00 MB in 0.36s
thread 7 random read totals (2770.62 MB/s) 1024.00 MB in 0.37s
thread 4 random read totals (2728.51 MB/s) 1024.00 MB in 0.38s
thread 1 random read totals (2720.31 MB/s) 1024.00 MB in 0.38s
thread 6 random read totals (2860.49 MB/s) 1024.00 MB in 0.36s
thread 2 random read totals (1153.87 MB/s) 1024.00 MB in 0.89s
thread 3 random read totals (1153.87 MB/s) 1024.00 MB in 0.89s
random read on /test/aiodio/junkfile (1.63 MB/s) 4.00 MB in 2.46s
thread 0 random read totals (1.62 MB/s) 4.00 MB in 2.47s
random read throughput (2905.81 MB/s) 7172.00 MB in 2.47s min transfer 4.00MB
real 0m5.748s
user 0m0.050s
sys 0m0.620s
[root@tst122 ~]#
RHEL4U2-32b domU guest:
-----------------------
[root@vs162 testcases]# time ./kernel/io/ltp-aiodio/aio-stress -I500 -o3 -O -r512 -t8 /test/aiodio/junkfile /test/aiodio/file2 /test/aiodio/file7 /test/aiodio/file8 /test/aiodio/file3 /test/aiodio/file4 /test/aiodio/file5 /test/aiodio/file6
adding stage random read
starting with random read
file size 1024MB, record size 512KB, depth 64, ios per iteration 8
max io_submit 64, buffer alignment set to 4KB
threads 8 files 8 contexts 1 context offset 2MB verification off
Running multi thread version num_threads:8
random read on /test/aiodio/file2 (7.30 MB/s) 1024.00 MB in 140.35s
thread 1 random read totals (7.27 MB/s) 1024.00 MB in 140.84s
random read on /test/aiodio/file4 (0.06 MB/s) 8.00 MB in 141.47s
thread 5 random read totals (0.06 MB/s) 8.00 MB in 141.60s
random read on /test/aiodio/file5 (0.03 MB/s) 4.00 MB in 141.83s
thread 6 random read totals (0.03 MB/s) 4.00 MB in 141.94s
random read on /test/aiodio/file8 (0.03 MB/s) 4.00 MB in 142.82s
thread 3 random read totals (0.03 MB/s) 4.00 MB in 142.92s
random read on /test/aiodio/file3 (0.03 MB/s) 4.00 MB in 143.04s
thread 4 random read totals (0.03 MB/s) 4.00 MB in 143.08s
random read on /test/aiodio/file6 (0.03 MB/s) 4.00 MB in 143.45s
thread 7 random read totals (0.03 MB/s) 4.00 MB in 143.52s
random read on /test/aiodio/file7 (0.03 MB/s) 4.00 MB in 143.80s
thread 2 random read totals (0.03 MB/s) 4.00 MB in 143.91s
random read on /test/aiodio/junkfile (0.03 MB/s) 4.00 MB in 145.83s
thread 0 random read totals (0.03 MB/s) 4.00 MB in 145.92s
random read throughput (7.24 MB/s) 1056.00 MB in 145.92s min transfer 4.00MB
real 2m31.215s
user 0m0.724s
sys 0m1.916s
[root@vs162 testcases]#
RHEL3U6 domU guest:
-------------------
[root@vs174 testcases]# time ./kernel/io/ltp-aiodio/aio-stress -I500 -o3 -O -r512 -t8 /test/aiodio/junkfile /test/aiodio/file2 /test/aiodio/file7 /test/aiodio/file8 /test/aiodio/file3 /test/aiodio/file4 /test/aiodio/file5 /test/aiodio/file6
adding stage random read
starting with random read
file size 1024MB, record size 512KB, depth 64, ios per iteration 8
max io_submit 64, buffer alignment set to 4KB
threads 8 files 8 contexts 1 context offset 2MB verification off
Running multi thread version num_threads:8
random read on /test/aiodio/file7 (0.10 MB/s) 1024.00 MB in 10322.46s
thread 2 random read totals (0.10 MB/s) 1024.00 MB in 10323.40s
random read on /test/aiodio/file4 (0.10 MB/s) 1024.00 MB in 10332.31s
thread 5 random read totals (0.10 MB/s) 1024.00 MB in 10332.59s
random read on /test/aiodio/file5 (0.10 MB/s) 1024.00 MB in 10344.51s
thread 6 random read totals (0.10 MB/s) 1024.00 MB in 10344.56s
random read on /test/aiodio/file6 (0.10 MB/s) 1024.00 MB in 10346.31s
thread 7 random read totals (0.10 MB/s) 1024.00 MB in 10346.88s
random read on /test/aiodio/file3 (0.10 MB/s) 1024.00 MB in 10353.64s
thread 4 random read totals (0.10 MB/s) 1024.00 MB in 10353.68s
random read on /test/aiodio/file8 (0.09 MB/s) 972.00 MB in 10373.37s
random read on /test/aiodio/file2 (0.08 MB/s) 860.00 MB in 10373.57s
random read on /test/aiodio/junkfile (0.08 MB/s) 860.00 MB in 10373.59s
thread 1 random read totals (0.08 MB/s) 860.00 MB in 10373.77s
thread 3 random read totals (0.09 MB/s) 972.00 MB in 10373.76s
thread 0 random read totals (0.08 MB/s) 860.00 MB in 10373.77s
random read throughput (0.75 MB/s) 7812.00 MB in 10373.77s min transfer 860.00MB
real 173m1.057s
user 0m1.140s
sys 0m49.050s
[root@vs174 testcases]#
[-- Attachment #4: Type: text/plain, Size: 138 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
next reply other threads:[~2006-05-10 19:51 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-10 19:51 Ed Smith [this message]
-- strict thread matches above, loose matches on Subject: below --
2006-05-06 18:21 Testing status of fully virtualized guests (Intel VT) on 32bit XEN unstable Ed Smith
2006-05-05 15:50 Ed Smith
2006-05-06 18:19 ` Nivedita Singhvi
2006-05-03 23:19 Ed Smith
2006-05-02 23:23 Ed Smith
2006-05-02 20:51 Ed Smith
2006-05-03 6:44 ` Keir Fraser
2006-05-01 23:33 Ed Smith
2006-04-30 1:45 Zhao, Yunfeng
2006-04-29 15:57 Ed Smith
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=44624441.6040609@virtualiron.com \
--to=esmith@virtualiron.com \
--cc=xen-devel@lists.xensource.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.