From: Daniel Hilst Selli <danielhilst@gmail.com>
To: "linux-c-programming@vger.kernel.org"
<linux-c-programming@vger.kernel.org>
Subject: Re: Show application performance/errors from pseudo file
Date: Thu, 23 Oct 2014 17:09:03 -0200 [thread overview]
Message-ID: <5449524F.1000206@gmail.com> (raw)
In-Reply-To: <54416747.2010108@gmail.com>
On 10/17/2014 04:00 PM, Daniel Hilst Selli wrote:
> On 10/17/2014 03:59 PM, Daniel Hilst Selli wrote:
>> On 10/16/2014 05:31 PM, Yichao Yu wrote:
>>> Resend as plain text.
>>>
>>> On Thu, Oct 16, 2014 at 4:01 PM, Yichao Yu <yyc1992@gmail.com> wrote:
>>>>
>>>>
>>>> On Thu, Oct 16, 2014 at 3:50 PM, Daniel Hilst Selli <danielhilst@gmail.com>
>>>> wrote:
>>>>>
>>>>> I'm writing a new application and would be nice to have a pseudo file
>>>>> showing its status, just the way that procfs does with kernel.
>>>>> I'm looking for sugestions, I want to `cat' files contents and have
>>>>> something similar to /proc/meminfo
>>>>>
>>>>> First I think using named pipes, but, AFAIK, pipes would retain data
>>>>> writed until someone read it, what I thought is a kind of read
>>>>> hook that only show data when asked for. Here are a few requisites,
>>>>>
>>>>> - Don't retain data
>>>>> - Don't generate disk I/O
>>>>> - Vanish when application stops
>>>>> - Work with a simple cat or something similar..
>>>>
>>>>
>>>> You should have a look at fuse[1].
>>>>
>>>> [1] http://fuse.sourceforge.net/
>>>>
>>>>>
>>>>>
>>>>> With that in mind I think about using unix domain sockets.. it seems to
>>>>> fit all requisites, for
>>>>> the fourth requisite I could use netcat, that is almost cat,
>>>>>
>>>>> Cheers
>>>>> --
>>>>> To unsubscribe from this list: send the line "unsubscribe
>>>>> linux-c-programming" in
>>>>> the body of a message to majordomo@vger.kernel.org
>>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>>>
>>>>
>>
>> Seems god, a little overkill for so simple stuff but adds nice features, I'm taking
>> a look on further options..
>>
>> Cheers
> good*
Looking further I found that named pipes will block on open() calls, until
the other end open it too. So I still can implement this with named pipes.
Here is the pseudo code implementation I use. (Yes I'm using threads)
void *stats_handler(void *unused)
{
for (;;) {
int fd = open(FIFO_PATH, O_WRONLY); /* should block here */
/* handle open error */
write(fd, statistics_buffer, statistics_buffer_len);
close(fd);
sleep(1); /* avoid flooders :-P */
}
}
void stats_init(void)
{
pthread_t stat_id;
mkfifo(FIFO_PATH, 0755);
/* handle mkfifo errors */
pthread_create(&stat_id, NULL, stats_handler, NULL);
}
int main(void)
{
...
stats_init();
...
}
After that I can just `cat FIFO_PATH' to see my statistics :D
Thanks for all and best regards!
prev parent reply other threads:[~2014-10-23 19:09 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-16 19:50 Show application performance/errors from pseudo file Daniel Hilst Selli
2014-10-16 20:00 ` mspiegelmock
[not found] ` <CAMvDr+T54gR3APzCOyZczvCTZs=FvHhsCA0VkZAVJR3BQdeT+w@mail.gmail.com>
2014-10-16 20:31 ` Yichao Yu
2014-10-17 18:59 ` Daniel Hilst Selli
2014-10-17 19:00 ` Daniel Hilst Selli
2014-10-23 19:09 ` Daniel Hilst Selli [this message]
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=5449524F.1000206@gmail.com \
--to=danielhilst@gmail.com \
--cc=linux-c-programming@vger.kernel.org \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).