James Manning [EMAIL PROTECTED] wrote:
[Gregory Leblanc]
[root@bod tiobench-0.3.1]# ./tiobench.pl --dir /raid5
No size specified, using 200 MB
Size is MB, BlkSz is Bytes, Read, Write, and Seeks are MB/sec
Try making the size at least double that of ram.
Actually, I do exactly
Bill BAO [EMAIL PROTECTED] wrote:
is there anybody doing the parity generation by using the
drive XOR cmd (XDWRITE, XDREAD, XDPWRITE) ?
And you know of many drives supporting this ?
(even the COPY command is no longer supported by any drives)
Pete Rossi [EMAIL PROTECTED] wrote:
That is.. looking to implement a fileserver that can handle any
single point of failure.. either a disk or processor.
You can share disks with the GFS project. Whether you can do RAID
at the same time today, I don't know.
Mike Bilow [EMAIL PROTECTED] wrote:
3. The reason your log stops is because your SCSI bus stops. If you have
another machine running syslogd, try pointing your log across the network;
see the section about "Remote Machine" in the "man syslog.conf" page.
When debugging bizarre problems, I
Klaus Schroer [EMAIL PROTECTED] wrote:
With a four disk setup you get a significant performance increase for
writing but not very much for reading (again software raid with one
AHA2940U2W LVD):
However, we see that each of the disks is at 10 MByte/s (still to be
verified), thus 4 disk is 3 x
Klaus Schroer [EMAIL PROTECTED] wrote:
This suggests that the DPT controller itself is not very efficient and
that even the Megaraid cannot beat a single controller software raid setup
(at least if you only use 3 disks). It might be a different stroy if you
Yes. However you will note that a
Gerrish, Robert [EMAIL PROTECTED] wrote:
backups can save your business. You not only should think about system
failures, but fires, floods, etc.
and silent data corruption. Remember, IDE hasn't parity. Also, disk-drives
can develop bad blocks over time.
That's why incremental backups, even
Matthew Clark [EMAIL PROTECTED] wrote:
I have a PIII 500 HP Netserver LH4 with 256Mb RAM
[ ... ]
If I run bonnie with 300Meg size, I get very low performance readings.
I would expect much higher throughoutput, especially with 256 MB
RAM. With 256 MB RAM, I usually performance-test on 2 GB
Tomas Fasth [EMAIL PROTECTED] wrote:
problem is that the newly inserted disk never get ordered to spin up.
Something that could be done (which used to be done in the scsi-idle
patch) was to detect NOT READY, POWER UP COMMAND REQUIRED (0x2 0x2A
or something like this) sense and to send a
Marc SCHAEFER [EMAIL PROTECTED] wrote:
I am going to try with raw IO (with sct patch) to see if it's the
memcpy_to_fs() (or 2.x equivalent) which is responsible for the slow down.
I am also going to try with two QLOGIC ISP1080 since they seem even faster
than the AIC7895 which was already
Marc SCHAEFER [EMAIL PROTECTED] wrote:
Now, RAID5 on the same 7 disk set:
---Sequential Output ---Sequential Input--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block---
MachineMB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU
I hope this is useful to someone. Basically, three SCSI adapters
(two LVD: AIC7895 and QLOGIC 1080, one Ultra: AIC789x), 8 cheatah
9 GB 10'000 RPM disks (each able to do around 18.5 MByte/s at the
start of the disk), 4 on the first LVD channel, two on each of the others.
440LX+ Intel motherboard
Francisco Jose Montilla [EMAIL PROTECTED] wrote:
- Doesn't SCSI controllers use parity? (Although you have to
enable it, of course)
Yes. However some of them do not use ``parity'' on the PCI bus, or in
their internal memory. This has happened to me (at a very stressing
case, with a bad
Dr. Michael Weller [EMAIL PROTECTED] wrote:
During the slow down, top claims system is well over 90% percent idle,
CPU time consumed by tar and general system time spent is virtually zero
Can you look in the Red Hat 5.0 errata list, and search for ``tar
slowdown'' ?
Eyal Lebedinsky [EMAIL PROTECTED] wrote:
I really think the disks are just stubborn.
Yes. Moreover they might handle disconnection unefficiently, which will
kill your RAID0 performance completely.
Nick [EMAIL PROTECTED] wrote:
Here's a bonnie result on 2 LVD Barracudas 4.5GB RAID-0 on a Symbios 896
ultra2 controller runnig Roudier's pre-sym8xx driver. (Dual P2 300Mhz
128MB RAM. 2.1.128)
You can't benchmark a 128MB machine with a 100 MB file.
Can you try with a 800 MB file ?
16 matches
Mail list logo