Here's a link to Windows documentation.
Basically I would like to get similar data, but on Linux. If not all is possible, then at least some parts.
If you enable CONFIG\_TASK\_IO\_ACCOUNTING, you will have the information available in /proc/<pid>/io. This is available since kernel 2.6.20, but not normally enabled by default (However, in Ubuntu 8.04 it seems like it is enabled).
You can read about the various data items in Documentation/filesystems/proc.txt in the kernel source tree. Especially section 2.14 should be of interest.
Have a look at /proc/PID/io - it's the current I/O accounting information of the process with PID.
Look at the pseudo-files under /proc/<PID>/. Maybe you can find what you need there.
Look at man 5 proc, or failing that the kernel documentation. However, I don't see much that looks promising. Sorry.
Perhaps you want getrusage()? Not all fields are maintained under linux however. Perhaps enabling the CONFIG_TASK_IO_ACCOUNTING will cause them to be maintained?
Related
I see that inspecting /proc/self/maps on Linux machines lets me see the pages that have been mapped in. As a result I can write a program to read and parse the pages it has mapped in.
How could one go about doing something similar for Windows? Are there any APIs for the same? If not, do you have any suggestions on how this could be done?
Yes, the possibility exists. First of all You have to access any process memory, or better, make it "accessible". Then You can read memory.
Here are some usefull links ( by the way, You should always look in there, if You come from linux and try to do things on windows, it is the main source ).
https://msdn.microsoft.com/en-us/library/windows/desktop/ms680553%28v=vs.85%29.aspx
https://msdn.microsoft.com/en-us/library/windows/desktop/aa366916%28v=vs.85%29.aspx
There is all documented.
But ther are also undocumented approaches, really crazy stuff, which also deals whith this topic.Like this for example.
http://undocumented.ntinternals.net/UserMode/Undocumented%20Functions/Memory%20Management/Virtual%20Memory/NtReadVirtualMemory.html
I have a new Ubuntu Linux Server 64bit 10.04 LTS.
A default install of Mysql with replication turned on appears to be leaking memory.
However, we've tried going back to an earlier version and memory is still leaking but I can't tell where.
What tools/techniques can I use to pinpoint where memory is leaking so that I can rectify the problem?
Valgrind, http://valgrind.org/, can be very useful in these situations. It runs on unmodified executables but it does help tremendously if you can install the debugging symbols. Be sure to use the --show-reachable=yes flag as the leaked memory may still be reachable in some way but just not the way you want it. Also --trace-children in case of a fork. You'll likely have to track down in the start-up script where the executable is called and then add something like the following:
valgrind --show-reachable=yes --trace-children=yes --log-file=/path/to/log SQL-cmdline sqlargs
The man page has lots of other potentially useful options.
Have you tried the MySQL mailing list? Something like this would certainly be of interest to them if you can reproduce it in a straightforward manner.
You can use Valgrind as ninjalj suggests, but I doubt you'll get that close to anything useful. Even if you see a real leak (and they will be hard enough to validate), tracking down the root cause through the C call stacks will likely be very annoying (for example if the leak is triggered by a particular SQL pattern or stored procedure, you'll be looking at the call stack from the resultant optimized query, and not the original calls, which are likely in a different language).
Normally you might have no recourse, and have to resort to tracking it down through callstacks and iterative testing, but you have the source code to MySQL (including the source for the exact default package install), so you can use more advanced tools like MemoryScape (or at least build with symbols in order to provide Valgrind more food for thought).
Try using valgrind.
A very good and powerful tool, which is installed/available for most distributions is Valgrind.
It has a plethora of different options and is pretty much (as far as I've seen) the default profiler under linux systems.
So a OS multi-tasks and runs one process at a given moment (assuming it's a single core machine). Is there a way I can track/poll and find which process was running at what time and at what time it was put in the queue and retrieved back. Are there any system calls for this ?. Is there a way to achieve this without modifying the linux kernel's source.
I think you need lttng, it definitely give a you a elaborate view of the system's task switch thing(and much more than that) with the lttng viewer. Lttng's kernel part has been merged to current Linux kernel, and you can use it if your kernel has enabled this feature. Here is some screen shots for lttng.
I don't think you can do this natively. AFAIK linux does not a keep a history track of this information.
That's an illogical question. If you are querying the OS from a script/process then the active program is ... YOURS.
Though I guess if you want the history you could watch the /proc directory or the output from ps
I am more common with the QNX os and I could change the arguments for the ide driver on startup. I wanted to change the arguments that get started in a linux kernel for the jfs filesystem which mounts a flash card. I can't seem to find a good web source for this information but I am sure I am not "googling" it right.
What are the different options available for the jfs/ide driver and where do I change them for the current running driver?
Thank you and I apologize this isn't exactly programming but this site always produces great answers when I ask a programming question so I figure it might help here too.
ok, i finally found out it is defined in the /etc/fstab file. Now I just need to find out what options are available and best for a compact flash card.
Options for most filesystems are shown in the "mount" manpage.
The ata driver may also have various tunables. You can probably change these in real time with "hdparm".
Some options might be filesystem create time options, in which case mkfs will document them.
See the relevant man pages.
If you want to pass options to jfs, you need to pass them as part of kernel boot
line, which is specified in your grub.conf or lilo.conf file. You can find the supported
options in the Documentation/ directory of the kernel source, or the jfs source itself.
As for the recommended mount options, you need to look at the mount(8) man page.
I would recommend noatime,nodirtime at a mininum. The default is atime, which
causes the file inode to be updated everytime you look at a file.
When I use top the iowait on the host is really high.
iostat tells me which disk is utilized more but I want to find out which process is the culprit?
I am trying to find this out on a red hat linux host. Any suggestions.
EDIT: My linux flavor does not either have atop or ntop and since building kernel is not an option for me don't ask me why :) (since this is not my personal box). are there any other alternatives
I usually use atop. There's a really good article at Debian Package A Day about it. It does require kernel patching (although Ubuntu already has the patch applied, I'm not sure about any other distributions.)
Use iotop.
Or you can get it standalone, it's a simple python script which requires a recent kernel (can't remember, but at least > 2.6.20)