there is the fastest way to load the data if there's a lot of it.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mys
atted with the FAT32
filesystem? FAT32 files are limited to 2^32-1 bytes - i.e. 1 less than
4GB.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To
On 18 May 2004, at 2:28 pm, RV Tec wrote:
Is MySQL able to handle such load with no problems/turbulences
at all? If so, what would be the best hardware/OS
configuration?
What is the largest DB known to MySQL community?
We regularly run databases with around 200 GB of data per instance,
MySQL 4.0.18, Tru64 5.1B
The log partition filled up on one of our instances. RESET MASTER
fails with:
ERROR 1186: Binlog closed, cannot RESET MASTER
I deleted some older logs by hand, but RESET MASTER still fails.
FLUSH LOGS appears to succeed, but no binary logs are being written,
and RESET M
Hmm, well, I seem to have resolved the problem, and it looks like it
was not MySQL's fault. I updated the OS to the current testing release
of Debian, which included a C library update. Following a reboot, the
code worked perfectly. Looks like this was an Itanium2 C library bug,
most likely.
On 14 May 2004, at 1:14 am, Dathan Vance Pattishall wrote:
-Original Message-
From: Jeremy Zawodny [mailto:[EMAIL PROTECTED]
Sent: Thursday, May 13, 2004 4:03 PM
To: Dathan Vance Pattishall
Cc: 'Tim Cutts'; 'MySQL List'
Subject: Re: InnoDB filesystem
On Thu, May 13
On 14 May 2004, at 4:37 am, Roy Butler wrote:
Jacob,
>> I'd go with Reiser on SuSE.
>
> What about Reiser on Debian?
I'd choose SuSE since Reiser is their default filesystem and they have
been an early implementor of Reiser-related patches. If you use Linux
kernel 2.4.24 (or later) and the lat
On 13 May 2004, at 4:02 pm, Jacob Friis Larsen wrote:
I'd go with Reiser on SuSE.
What about Reiser on Debian?
It shouldn't matter too much. This functionality is in the kernel, so
if the kernel version on SuSE and Debian is the same, the filesystem
code will be the same, with the possible cav
On 13 May 2004, at 3:34 pm, Dan Nelson wrote:
Pros: performance and bypassing the filesystem cache.
I believe most OSes support direct file access which either bypasses or
minimizes cache effects, and InnoDB will enable it if possible.
Solaris direct file I/O performance on UFS is within a couple
Dear MySQL experts!
I'm at something of a loss here. I'm testing MySQL on a new hardware
platform. Previously, we had it running on Tru64 Alpha boxes. We're
now moving it onto Itanium2 boxes running Debian Linux. Each machine
has 4 CPUs and 16 GB RAM. Kernel version is 2.6.5.
We couldn't
?
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
7;t have in its own buffers but
which nevertheless are frequently accessed will already be in RAM, and
therefore faster to access.
Even so, you'd probably do better with a 64 bit processor with that
amount of memory.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sang
On 30 Mar 2004, at 09:05, Tim Cutts wrote:
SATA RAID devices aren't that bad, you know, and they are a lot
cheaper than equivalent amounts of SCSI storage. We've used NexSan
ATABoy devices, which are relatively cheap, and get you a lot of
storage in very little space (10GB in a 3U b
ySQL servers disks are 15K RPM
FibreChannel disks on HP StorageWorks HSV110 controllers, which is
rather more at the upper end of the scale. ;-)
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL General Mailing List
For list archives
very connection
thread if the server supports transactional table types.
As a followup question; what happens to the binlog cache if a thread
requires more? Does it automatically increase it as needed (up to an
eventual limit of max_binlog_cache_size)?
Tim
--
Dr Tim Cutts
Informatics Systems
the database before the dump is done?
mysqldump, if used with the --lock-tables option (which is implied by
--opt) obtains read locks, so queries attempting to update the database
will block until the dump has finished.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger
On 25 Mar 2004, at 10:10, Tim Cutts wrote:
No, indeed. I'm going to try building mysql myself, on the machine on
which it's going to be running, and see whether that still has the
issue...
The version compiled natively on the machine does the same thing
(although it uses a little l
/InnoDB bug. I have
not
heard of a similar memory problem from anyone else.
No, indeed. I'm going to try building mysql myself, on the machine on
which it's going to be running, and see whether that still has the
issue...
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Tr
t --skip-innodb just changes some memory
allocation patters on startup, which possibly avoid triggering the
bug.
Perhaps...
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL General Mailing List
For list archives: http://list
On 22 Mar 2004, at 18:24, Tim Cutts wrote:
Some users' code is causing MySQL's memory use to explode. By the
time we reach about 200 simultaneous connections, the MySQL server is
using 8GB of virtual memory, and then falls over (the machine is an
AlphaServer ES45 with 8 GB of physi
ed perl scripts on perl built with icc 7.1 and gcc 3.2, and the
performance difference was negligible.
I wouldn't bust a gut over this, if I were you. I suspect MySQL is not
CPU-bound, most of the time, and certainly not floating-point bound
(which is where icc really shines)
Tim
--
Dr Ti
It's my understanding that MySQL does not currently compile with the
Intel compiler - it's too gcc-specific, and icc is not 100%
gcc-compatible. I seem to remember seeing this in the on-line
documentation somewhere, but I can't remember where.
Tim
--
Dr Tim Cutts
Informatic
ent device
from the databases.
As an additional point, if I add a replication slave to the equation,
the time becomes:
2518 seconds
Which is scarcely any difference at all.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB10 1SA, UK
--
MySQL Gene
y ideas, including ways I can get MySQL to tell me more about what
it's doing, would be most helpful. The query log, even with
log-warnings on, does not tell us much.
Many thanks in advance...
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Ca
On 16 Mar 2004, at 22:30, Gabriel Guzman wrote:
On Tuesday 16 March 2004 02:19 pm, Tim Cutts wrote:
This is something the list admins might be able to sort out. It's
common for mailing lists to set a reply-to header...
here we go again
Oops - is that a bad button to push? :-)
If
be able to sort out. It's
common for mailing lists to set a reply-to header, so that replies
automatically go to the list, and not to the originating poster (all
the mailing lists that I run do so) but this list does not.
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust S
commercial utilities exist for backups of MySQL instances? For
complicated reasons I don't want to go into, most of the "standard"
methods for backing up MySQL instances don't work well for us, or at
least have significant drawbacks.
Tim
--
Dr Tim Cutts
Informatics Systems Group
W
es (and so has another user).
The software in use on both client machines is identical, and the user
accounts are on shared home directories, and so would have been reading
the same ~/.my.cnf
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambridge, CB
eally irritating if you're debugging a problem with a busy
production server. You don't want the general query log on all the
time, but you don't want to have to have database downtime (even if
it's only momentary) to switch the logging on and then off again.
Tim
--
Dr Tim Cutts
assword: YES)
I've tried re-starting the database, to no effect (fortunately this is
just a test instance).
Am I missing something ridiculously simple, or is this a bug?
Thanks in advance...
Tim
--
Dr Tim Cutts
Informatics Systems Group
Wellcome Trust Sanger Institute
Hinxton, Cambrid
30 matches
Mail list logo