path: root/src/sys_fs_cgroup.c

#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include <stdio.h>
#include <stdlib.h>
#include <inttypes.h>
#include <sys/types.h>
#include <dirent.h>
#include <string.h>

#include "common.h"
#include "appconfig.h"
#include "procfile.h"
#include "log.h"
#include "rrd.h"
#include "main.h"
#include "popen.h"
#include "proc_self_mountinfo.h"

// ----------------------------------------------------------------------------
// cgroup globals

static int cgroup_enable_cpuacct_stat = CONFIG_ONDEMAND_ONDEMAND;
static int cgroup_enable_cpuacct_usage = CONFIG_ONDEMAND_ONDEMAND;
static int cgroup_enable_memory = CONFIG_ONDEMAND_ONDEMAND;
static int cgroup_enable_blkio = CONFIG_ONDEMAND_ONDEMAND;
static int cgroup_enable_new_cgroups_detected_at_runtime = 1;
static int cgroup_check_for_new_every = 10;
static char *cgroup_cpuacct_base = NULL;
static char *cgroup_blkio_base = NULL;
static char *cgroup_memory_base = NULL;

static int cgroup_root_count = 0;
static int cgroup_root_max = 500;
static int cgroup_max_depth = 0;

void read_cgroup_plugin_configuration() {
	cgroup_check_for_new_every = config_get_number("plugin:cgroups", "check for new cgroups every", cgroup_check_for_new_every);

	cgroup_enable_cpuacct_stat = config_get_boolean_ondemand("plugin:cgroups", "enable cpuacct stat", cgroup_enable_cpuacct_stat);
	cgroup_enable_cpuacct_usage = config_get_boolean_ondemand("plugin:cgroups", "enable cpuacct usage", cgroup_enable_cpuacct_usage);
	cgroup_enable_memory = config_get_boolean_ondemand("plugin:cgroups", "enable memory", cgroup_enable_memory);
	cgroup_enable_blkio = config_get_boolean_ondemand("plugin:cgroups", "enable blkio", cgroup_enable_blkio);

	char filename[FILENAME_MAX + 1], *s;
	struct mountinfo *mi, *root = mountinfo_read();

	mi = mountinfo_find_by_filesystem_mount_source(root, "cgroup", "cpuacct");
	if(!mi) mi = mountinfo_find_by_filesystem_super_option(root, "cgroup", "cpuacct");
	if(!mi) s = "/sys/fs/cgroup/cpuacct";
	else s = mi->mount_point;
	snprintfz(filename, FILENAME_MAX, "%s%s", global_host_prefix, s);
	cgroup_cpuacct_base = config_get("plugin:cgroups", "path to /sys/fs/cgroup/cpuacct", filename);

	mi = mountinfo_find_by_filesystem_mount_source(root, "cgroup", "blkio");
	if(!mi) mi = mountinfo_find_by_filesystem_super_option(root, "cgroup", "blkio");
	if(!mi) s = "/sys/fs/cgroup/blkio";
	else s = mi->mount_point;
	snprintfz(filename, FILENAME_MAX, "%s%s", global_host_prefix, s);
	cgroup_blkio_base = config_get("plugin:cgroups", "path to /sys/fs/cgroup/blkio", filename);

	mi = mountinfo_find_by_filesystem_mount_source(root, "cgroup", "memory");
	if(!mi) mi = mountinfo_find_by_filesystem_super_option(root, "cgroup", "memory");
	if(!mi) s = "/sys/fs/cgroup/memory";
	else s = mi->mount_point;
	snprintfz(filename, FILENAME_MAX, "%s%s", global_host_prefix, s);
	cgroup_memory_base = config_get("plugin:cgroups", "path to /sys/fs/cgroup/memory", filename);

	cgroup_root_max = config_get_number("plugin:cgroups", "max cgroups to allow", cgroup_root_max);
	cgroup_max_depth = config_get_number("plugin:cgroups", "max cgroups depth to monitor", cgroup_max_depth);

	cgroup_enable_new_cgroups_detected_at_runtime = config_get_boolean("plugin:cgroups", "enable new cgroups detected at run time", cgroup_enable_new_cgroups_detected_at_runtime);

	mountinfo_free(root);
}

// ----------------------------------------------------------------------------
// cgroup objects

struct blkio {
	int updated;

	char *filename;

	unsigned long long Read;
	unsigned long long Write;
/*
	unsigned long long Sync;
	unsigned long long Async;
	unsigned long long Total;
*/
};

// https://www.kernel.org/doc/Documentation/cgroup-v1/memory.txt
struct memory {
	int updated;

	char *filename;

	int has_dirty_swap;

	unsigned long long cache;
	unsigned long long rss;
	unsigned long long rss_huge;
	unsigned long long mapped_file;
	unsigned long long writeback;
	unsigned long long dirty;
	unsigned long long swap;
	unsigned long long pgpgin;
	unsigned long long pgpgout;
	unsigned long long pgfault;
	unsigned long long pgmajfault;
/*
	unsigned long long inactive_anon;
	unsigned long long active_anon;
	unsigned long long inactive_file;
	unsigned long long active_file;
	unsigned long long unevictable;
	unsigned long long hierarchical_memory_limit;
	unsigned long long total_cache;
	unsigned long long total_rss;
	unsigned long long total_rss_huge;
	unsigned long long total_mapped_file;
	unsigned long long total_writeback;
	unsigned long long total_dirty;
	unsigned long long total_swap;
	unsigned long long total_pgpgin;
	unsigned long long total_pgpgout;
	unsigned long long total_pgfault;
	unsigned long long total_pgmajfault;
	unsigned long long total_inactive_anon;
	unsigned long long total_active_anon;
	unsigned long long total_inactive_file;
	unsigned long long total_active_file;
	unsigned long long total_unevictable;
*/
};

// https://www.kernel.org/doc/Documentation/cgroup-v1/cpuacct.txt
struct cpuacct_stat {
	int updated;

	char *filename;

	unsigned long long user;
	unsigned long long system;
};

// https://www.kernel.org/doc/Documentation/cgroup-v1/cpuacct.txt
struct cpuacct_usage {
	int updated;

	char *filename;

	unsigned int cpus;
	unsigned long long *cpu_percpu;
};

struct cgroup {
	int available;		// found in the filesystem
	int enabled;		// enabled in the config

	char *id;
	uint32_t hash;

	char *chart_id;
	char *chart_title;

	struct cpuacct_stat cpuacct_stat;
	struct cpuacct_usage cpuacct_usage;

	struct memory memory;

	struct blkio io_service_bytes;				// bytes
	struct blkio io_serviced;					// operations

	struct blkio throttle_io_service_bytes;		// bytes
	struct blkio throttle_io_serviced;			// operations

	struct blkio io_merged;						// operations
	struct blkio io_queued;						// operations

	struct cgroup *next;

} *cgroup_root = NULL;

// ----------------------------------------------------------------------------
// read values from /sys

void cgroup_read_cpuacct_stat(struct cpuacct_stat *cp) {
	static procfile *ff = NULL;

	static uint32_t user_hash = 0;
	static uint32_t system_hash = 0;

	if(unlikely(user_hash == 0)) {
		user_hash = simple_hash("user");
		system_hash = simple_hash("system");
	}

	cp->updated = 0;
	if(cp->filename) {
		ff = procfile_reopen(ff, cp->filename, NULL, PROCFILE_FLAG_DEFAULT);
		if(!ff) return;

		ff = procfile_readall(ff);
		if(!ff) return;

		unsigned long i, lines = procfile_lines(ff);

		if(lines < 1) {
			error("File '%s' should have 1+ lines.", cp->filename);
			return;
		}

		for(i = 0; i < lines ; i++) {
			char *s = procfile_lineword(ff, i, 0);
			uint32_t hash = simple_hash(s);

			if(hash == user_hash && !strcmp(s, "user"))
				cp->user = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == system_hash && !strcmp(s, "system"))
				cp->system = strtoull(procfile_lineword(ff, i, 1), NULL, 10);
		}

		cp->updated = 1;

		// fprintf(stderr, "READ '%s': user: %llu, system: %llu\n", cp->filename, cp->user, cp->system);
	}
}

void cgroup_read_cpuacct_usage(struct cpuacct_usage *ca) {
	static procfile *ff = NULL;

	ca->updated = 0;
	if(ca->filename) {
		ff = procfile_reopen(ff, ca->filename, NULL, PROCFILE_FLAG_DEFAULT);
		if(!ff) return;

		ff = procfile_readall(ff);
		if(!ff) return;

		if(procfile_lines(ff) < 1) {
			error("File '%s' should have 1+ lines but has %d.", ca->filename, procfile_lines(ff));
			return;
		}

		unsigned long i = procfile_linewords(ff, 0);
		if(i <= 0) return;

		// we may have 1 more CPU reported
		while(i > 0) {
			char *s = procfile_lineword(ff, 0, i - 1);
			if(!*s) i--;
			else break;
		}

		if(i != ca->cpus) {
			free(ca->cpu_percpu);

			ca->cpu_percpu = malloc(sizeof(unsigned long long) * i);
			if(!ca->cpu_percpu)
				fatal("Cannot allocate memory (%z bytes)", sizeof(unsigned long long) * i);

			ca->cpus = i;
		}

		for(i = 0; i < ca->cpus ;i++) {
			ca->cpu_percpu[i] = strtoull(procfile_lineword(ff, 0, i), NULL, 10);
			// fprintf(stderr, "READ '%s': cpu%d/%d: %llu ('%s')\n", ca->filename, i, ca->cpus, ca->cpu_percpu[i], procfile_lineword(ff, 0, i));
		}

		ca->updated = 1;
	}
}

void cgroup_read_blkio(struct blkio *io) {
	static procfile *ff = NULL;

	static uint32_t Read_hash = 0;
	static uint32_t Write_hash = 0;
/*
	static uint32_t Sync_hash = 0;
	static uint32_t Async_hash = 0;
	static uint32_t Total_hash = 0;
*/

	if(unlikely(Read_hash == 0)) {
		Read_hash = simple_hash("Read");
		Write_hash = simple_hash("Write");
/*
		Sync_hash = simple_hash("Sync");
		Async_hash = simple_hash("Async");
		Total_hash = simple_hash("Total");
*/
	}

	io->updated = 0;
	if(io->filename) {
		ff = procfile_reopen(ff, io->filename, NULL, PROCFILE_FLAG_DEFAULT);
		if(!ff) return;

		ff = procfile_readall(ff);
		if(!ff) return;

		unsigned long i, lines = procfile_lines(ff);

		if(lines < 1) {
			error("File '%s' should have 1+ lines.", io->filename);
			return;
		}

		io->Read = 0;
		io->Write = 0;
/*
		io->Sync = 0;
		io->Async = 0;
		io->Total = 0;
*/

		for(i = 0; i < lines ; i++) {
			char *s = procfile_lineword(ff, i, 1);
			uint32_t hash = simple_hash(s);

			if(hash == Read_hash && !strcmp(s, "Read"))
				io->Read += strtoull(procfile_lineword(ff, i, 2), NULL, 10);

			else if(hash == Write_hash && !strcmp(s, "Write"))
				io->Write += strtoull(procfile_lineword(ff, i, 2), NULL, 10);

/*
			else if(hash == Sync_hash && !strcmp(s, "Sync"))
				io->Sync += strtoull(procfile_lineword(ff, i, 2), NULL, 10);

			else if(hash == Async_hash && !strcmp(s, "Async"))
				io->Async += strtoull(procfile_lineword(ff, i, 2), NULL, 10);

			else if(hash == Total_hash && !strcmp(s, "Total"))
				io->Total += strtoull(procfile_lineword(ff, i, 2), NULL, 10);
*/
		}

		io->updated = 1;
		// fprintf(stderr, "READ '%s': Read: %llu, Write: %llu, Sync: %llu, Async: %llu, Total: %llu\n", io->filename, io->Read, io->Write, io->Sync, io->Async, io->Total);
	}
}

void cgroup_read_memory(struct memory *mem) {
	static procfile *ff = NULL;

	static uint32_t cache_hash = 0;
	static uint32_t rss_hash = 0;
	static uint32_t rss_huge_hash = 0;
	static uint32_t mapped_file_hash = 0;
	static uint32_t writeback_hash = 0;
	static uint32_t dirty_hash = 0;
	static uint32_t swap_hash = 0;
	static uint32_t pgpgin_hash = 0;
	static uint32_t pgpgout_hash = 0;
	static uint32_t pgfault_hash = 0;
	static uint32_t pgmajfault_hash = 0;
/*
	static uint32_t inactive_anon_hash = 0;
	static uint32_t active_anon_hash = 0;
	static uint32_t inactive_file_hash = 0;
	static uint32_t active_file_hash = 0;
	static uint32_t unevictable_hash = 0;
	static uint32_t hierarchical_memory_limit_hash = 0;
	static uint32_t total_cache_hash = 0;
	static uint32_t total_rss_hash = 0;
	static uint32_t total_rss_huge_hash = 0;
	static uint32_t total_mapped_file_hash = 0;
	static uint32_t total_writeback_hash = 0;
	static uint32_t total_dirty_hash = 0;
	static uint32_t total_swap_hash = 0;
	static uint32_t total_pgpgin_hash = 0;
	static uint32_t total_pgpgout_hash = 0;
	static uint32_t total_pgfault_hash = 0;
	static uint32_t total_pgmajfault_hash = 0;
	static uint32_t total_inactive_anon_hash = 0;
	static uint32_t total_active_anon_hash = 0;
	static uint32_t total_inactive_file_hash = 0;
	static uint32_t total_active_file_hash = 0;
	static uint32_t total_unevictable_hash = 0;
*/
	if(unlikely(cache_hash == 0)) {
		cache_hash = simple_hash("cache");
		rss_hash = simple_hash("rss");
		rss_huge_hash = simple_hash("rss_huge");
		mapped_file_hash = simple_hash("mapped_file");
		writeback_hash = simple_hash("writeback");
		dirty_hash = simple_hash("dirty");
		swap_hash = simple_hash("swap");
		pgpgin_hash = simple_hash("pgpgin");
		pgpgout_hash = simple_hash("pgpgout");
		pgfault_hash = simple_hash("pgfault");
		pgmajfault_hash = simple_hash("pgmajfault");
/*
		inactive_anon_hash = simple_hash("inactive_anon");
		active_anon_hash = simple_hash("active_anon");
		inactive_file_hash = simple_hash("inactive_file");
		active_file_hash = simple_hash("active_file");
		unevictable_hash = simple_hash("unevictable");
		hierarchical_memory_limit_hash = simple_hash("hierarchical_memory_limit");
		total_cache_hash = simple_hash("total_cache");
		total_rss_hash = simple_hash("total_rss");
		total_rss_huge_hash = simple_hash("total_rss_huge");
		total_mapped_file_hash = simple_hash("total_mapped_file");
		total_writeback_hash = simple_hash("total_writeback");
		total_dirty_hash = simple_hash("total_dirty");
		total_swap_hash = simple_hash("total_swap");
		total_pgpgin_hash = simple_hash("total_pgpgin");
		total_pgpgout_hash = simple_hash("total_pgpgout");
		total_pgfault_hash = simple_hash("total_pgfault");
		total_pgmajfault_hash = simple_hash("total_pgmajfault");
		total_inactive_anon_hash = simple_hash("total_inactive_anon");
		total_active_anon_hash = simple_hash("total_active_anon");
		total_inactive_file_hash = simple_hash("total_inactive_file");
		total_active_file_hash = simple_hash("total_active_file");
		total_unevictable_hash = simple_hash("total_unevictable");
*/
	}

	mem->updated = 0;
	if(mem->filename) {
		ff = procfile_reopen(ff, mem->filename, NULL, PROCFILE_FLAG_DEFAULT);
		if(!ff) return;

		ff = procfile_readall(ff);
		if(!ff) return;

		unsigned long i, lines = procfile_lines(ff);

		if(lines < 1) {
			error("File '%s' should have 1+ lines.", mem->filename);
			return;
		}

		for(i = 0; i < lines ; i++) {
			char *s = procfile_lineword(ff, i, 0);
			uint32_t hash = simple_hash(s);

			if(hash == cache_hash && !strcmp(s, "cache"))
				mem->cache = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == rss_hash && !strcmp(s, "rss"))
				mem->rss = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == rss_huge_hash && !strcmp(s, "rss_huge"))
				mem->rss_huge = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == mapped_file_hash && !strcmp(s, "mapped_file"))
				mem->mapped_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == writeback_hash && !strcmp(s, "writeback"))
				mem->writeback = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == dirty_hash && !strcmp(s, "dirty")) {
				mem->dirty = strtoull(procfile_lineword(ff, i, 1), NULL, 10);
				mem->has_dirty_swap = 1;
			}

			else if(hash == swap_hash && !strcmp(s, "swap")) {
				mem->swap = strtoull(procfile_lineword(ff, i, 1), NULL, 10);
				mem->has_dirty_swap = 1;
			}

			else if(hash == pgpgin_hash && !strcmp(s, "pgpgin"))
				mem->pgpgin = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == pgpgout_hash && !strcmp(s, "pgpgout"))
				mem->pgpgout = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == pgfault_hash && !strcmp(s, "pgfault"))
				mem->pgfault = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == pgmajfault_hash && !strcmp(s, "pgmajfault"))
				mem->pgmajfault = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

/*
			else if(hash == inactive_anon_hash && !strcmp(s, "inactive_anon"))
				mem->inactive_anon = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == active_anon_hash && !strcmp(s, "active_anon"))
				mem->active_anon = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == inactive_file_hash && !strcmp(s, "inactive_file"))
				mem->inactive_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == active_file_hash && !strcmp(s, "active_file"))
				mem->active_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == unevictable_hash && !strcmp(s, "unevictable"))
				mem->unevictable = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == hierarchical_memory_limit_hash && !strcmp(s, "hierarchical_memory_limit"))
				mem->hierarchical_memory_limit = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

	    	else if(hash == total_cache_hash && !strcmp(s, "total_cache"))
				mem->total_cache = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_rss_hash && !strcmp(s, "total_rss"))
				mem->total_rss = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_rss_huge_hash && !strcmp(s, "total_rss_huge"))
				mem->total_rss_huge = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_mapped_file_hash && !strcmp(s, "total_mapped_file"))
				mem->total_mapped_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_writeback_hash && !strcmp(s, "total_writeback"))
				mem->total_writeback = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_dirty_hash && !strcmp(s, "total_dirty"))
				mem->total_dirty = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_swap_hash && !strcmp(s, "total_swap"))
				mem->total_swap = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_pgpgin_hash && !strcmp(s, "total_pgpgin"))
				mem->total_pgpgin = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_pgpgout_hash && !strcmp(s, "total_pgpgout"))
				mem->total_pgpgout = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_pgfault_hash && !strcmp(s, "total_pgfault"))
				mem->total_pgfault = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_pgmajfault_hash && !strcmp(s, "total_pgmajfault"))
				mem->total_pgmajfault = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_inactive_anon_hash && !strcmp(s, "total_inactive_anon"))
				mem->total_inactive_anon = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_active_anon_hash && !strcmp(s, "total_active_anon"))
				mem->total_active_anon = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_inactive_file_hash && !strcmp(s, "total_inactive_file"))
				mem->total_inactive_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_active_file_hash && !strcmp(s, "total_active_file"))
				mem->total_active_file = strtoull(procfile_lineword(ff, i, 1), NULL, 10);

			else if(hash == total_unevictable_hash && !strcmp(s, "total_unevictable"))
				mem->total_unevictable = strtoull(procfile_lineword(ff, i, 1), NULL, 10);
*/
		}

		// fprintf(stderr, "READ: '%s', cache: %llu, rss: %llu, rss_huge: %llu, mapped_file: %llu, writeback: %llu, dirty: %llu, swap: %llu, pgpgin: %llu, pgpgout: %llu, pgfault: %llu, pgmajfault: %llu, inactive_anon: %llu, active_anon: %llu, inactive_file: %llu, active_file: %llu, unevictable: %llu, hierarchical_memory_limit: %llu, total_cache: %llu, total_rss: %llu, total_rss_huge: %llu, total_mapped_file: %llu, total_writeback: %llu, total_dirty: %llu, total_swap: %llu, total_pgpgin: %llu, total_pgpgout: %llu, total_pgfault: %llu, total_pgmajfault: %llu, total_inactive_anon: %llu, total_active_anon: %llu, total_inactive_file: %llu, total_active_file: %llu, total_unevictable: %llu\n", mem->filename, mem->cache, mem->rss, mem->rss_huge, mem->mapped_file, mem->writeback, mem->dirty, mem->swap, mem->pgpgin, mem->pgpgout, mem->pgfault, mem->pgmajfault, mem->inactive_anon, mem->active_anon, mem->inactive_file, mem->active_file, mem->unevictable, mem->hierarchical_memory_limit, mem->total_cache, mem->total_rss, mem->total_rss_huge, mem->total_mapped_file, mem->total_writeback, mem->total_dirty, mem->total_swap, mem->total_pgpgin, mem->total_pgpgout, mem->total_pgfault, mem->total_pgmajfault, mem->total_inactive_anon, mem->total_active_anon, mem->total_inactive_file, mem->total_active_file, mem->total_unevictable);

		mem->updated = 1;
	}
}

void cgroup_read(struct cgroup *cg) {
	debug(D_CGROUP, "reading metrics for cgroups '%s'", cg->id);

	cgroup_read_cpuacct_stat(&cg->cpuacct_stat);
	cgroup_read_cpuacct_usage(&cg->cpuacct_usage);
	cgroup_read_memory(&cg->memory);
	cgroup_read_blkio(&cg->io_service_bytes);
	cgroup_read_blkio(&cg->io_serviced);
	cgroup_read_blkio(&cg->throttle_io_service_bytes);
	cgroup_read_blkio(&cg->throttle_io_serviced);
	cgroup_read_blkio(&cg->io_merged);
	cgroup_read_blkio(&cg->io_queued);
}

void read_all_cgroups(struct cgroup *root) {
	debug(D_CGROUP, "reading metrics for all cgroups");

	struct cgroup *cg;

	for(cg = root; cg ; cg = cg->next)
		if(cg->enabled && cg->available)
			cgroup_read(cg);
}

// ----------------------------------------------------------------------------
// add/remove/find cgroup objects

#define CGROUP_CHARTID_LINE_MAX 1024

void cgroup_get_chart_id(struct cgroup *cg) {
	debug(D_CGROUP, "getting the name of cgroup '%s'", cg->id);

	pid_t cgroup_pid;
	char buffer[CGROUP_CHARTID_LINE_MAX + 1];

	snprintfz(buffer, CGROUP_CHARTID_LINE_MAX, "exec %s '%s'",
	         config_get("plugin:cgroups", "script to get cgroup names", PLUGINS_DIR "/cgroup-name.sh"), cg->chart_id);

	debug(D_CGROUP, "executing command '%s' for cgroup '%s'", buffer, cg->id);
	FILE *fp = mypopen(buffer, &cgroup_pid);
	if(!fp) {
		error("CGROUP: Cannot popen(\"%s\", \"r\").", buffer);
		return;
	}
	debug(D_CGROUP, "reading from command '%s' for cgroup '%s'", buffer, cg->id);
	char *s = fgets(buffer, CGROUP_CHARTID_LINE_MAX, fp);
	debug(D_CGROUP, "closing command for cgroup '%s'", cg->id);
	mypclose(fp, cgroup_pid);
	debug(D_CGROUP, "closed command for cgroup '%s'", cg->id);

	if(s && *s && *s != '\n') {
		debug(D_CGROUP, "cgroup '%s' should be renamed to '%s'", cg->id, s);

		trim(s);

		free(cg->chart_title);
		cg->chart_title = strdup(s);
		if(!cg->chart_title)
			fatal("CGROUP: Cannot allocate memory for chart name of cgroup '%s' chart name: '%s'", cg->id, s);

		netdata_fix_chart_name(cg->chart_title);

		free(cg->chart_id);
		cg->chart_id = strdup(s);
		if(!cg->chart_id)
			fatal("CGROUP: Cannot allocate memory for chart id of cgroup '%s' chart id: '%s'", cg->id, s);

		netdata_fix_chart_id(cg->chart_id);

		debug(D_CGROUP, "cgroup '%s' renamed to '%s' (title: '%s')", cg->id, cg->chart_id, cg->chart_title);
	}
	else debug(D_CGROUP, "cgroup '%s' is not to be renamed (will be shown as '%s')", cg->id, cg->chart_id);
}

struct cgroup *cgroup_add(const char *id) {
	debug(D_CGROUP, "adding cgroup '%s'", id);

	if(cgroup_root_count >= cgroup_root_max) {
		info("Maximum number of cgroups reached (%d). Not adding cgroup '%s'", cgroup_root_count, id);
		return NULL;
	}

	int def = cgroup_enable_new_cgroups_detected_at_runtime;
	const char *chart_id = id;
	if(!*chart_id) {
		chart_id = "/";

		// disable by default the root cgroup
		def = 0;
		debug(D_CGROUP, "cgroup '%s' is the root container (by default %s)", id, (def)?"enabled":"disabled");
	}
	else {
		if(*chart_id == '/') chart_id++;

		size_t len = strlen(chart_id);

		// disable by default the parent cgroup
		// for known cgroup managers
		if(!strcmp(chart_id, "lxc") ||
				!strcmp(chart_id, "docker") ||
				!strcmp(chart_id, "libvirt") ||
				!strcmp(chart_id, "qemu") ||
				!strcmp(chart_id, "systemd") ||
				!strcmp(chart_id, "system.slice") ||
				!strcmp(chart_id, "machine.slice") ||
				!strcmp(chart_id, "user") ||
				!strcmp(chart_id, "system") ||
				!strcmp(chart_id, "machine") ||
		   		// starts with them
				(len >  6 && !strncmp(chart_id, "user/", 6)) ||
				(len > 11 && !strncmp(chart_id, "user.slice/", 11)) ||
				// ends with them
				(len >  5 && !strncmp(&chart_id[len -  5], ".user", 5)) ||
				(len >  5 && !strncmp(&chart_id[len -  5], ".swap", 5)) ||
				(len >  6 && !strncmp(&chart_id[len -  6], ".slice", 6)) ||
				(len >  6 && !strncmp(&chart_id[len -  6], ".mount", 6)) ||
				(len >  8 && !strncmp(&chart_id[len -  8], ".session", 8)) ||
				(len >  8 && !strncmp(&chart_id[len -  8], ".service", 8)) ||
				(len > 10 && !strncmp(&chart_id[len - 10], ".partition", 10))
				) {
			def = 0;
			debug(D_CGROUP, "cgroup '%s' is %s (by default)", id, (def)?"enabled":"disabled");
		}
	}

	struct cgroup *cg = calloc(1, sizeof(struct cgroup));
	if(!cg) fatal("Cannot allocate memory for cgroup '%s'", id);

	debug(D_CGROUP, "adding cgroup '%s'", id);

	cg->id = strdup(id);
	if(!cg->id) fatal("Cannot allocate memory for cgroup '%s'", id);

	cg->hash = simple_hash(cg->id);

	cg->chart_id = strdup(chart_id);
	if(!cg->chart_id) fatal("Cannot allocate memory for cgroup '%s'", id);

	cg->chart_title = strdup(chart_id);
	if(!cg->chart_title) fatal("Cannot allocate memory for cgroup '%s'", id);

	if(!cgroup_root)
		cgroup_root = cg;
	else {
		// append it
		struct cgroup *e;
		for(e = cgroup_root; e->next ;e = e->next) ;
		e->next = cg;
	}

	cgroup_root_count++;

	// fix the name by calling the external script
	cgroup_get_chart_id(cg);

	char option[FILENAME_MAX + 1];
	snprintfz(option, FILENAME_MAX, "enable cgroup %s", cg->chart_title);
	cg->enabled = config_get_boolean("plugin:cgroups", option, def);

	debug(D_CGROUP, "Added cgroup '%s' with chart id '%s' and title '%s' as %s (default was %s)", cg->id, cg->chart_id, cg->chart_title, (cg->enabled)?"enabled":"disabled", (def)?"enabled":"disabled");

	return cg;
}

void cgroup_free(struct cgroup *cg) {
	debug(D_CGROUP, "Removing cgroup '%s' with chart id '%s' (was %s and %s)", cg->id, cg->chart_id, (cg->enabled)?"enabled":"disabled", (cg->available)?"available":"not available");

	free(cg->cpuacct_usage.cpu_percpu);

	free(cg->cpuacct_stat.filename);
	free(cg->cpuacct_usage.filename);
	free(cg->memory.filename);
	free(cg->io_service_bytes.filename);
	free(cg->io_serviced.filename);
	free(cg->throttle_io_service_bytes.filename);
	free(cg->throttle_io_serviced.filename);
	free(cg->io_merged.filename);
	free(cg->io_queued.filename);

	free(cg->id);
	free(cg->chart_id);
	free(cg->chart_title);
	free(cg);

	cgroup_root_count--;
}

// find if a given cgroup exists
struct cgroup *cgroup_find(const char *id) {
	debug(D_CGROUP, "searching for cgroup '%s'", id);

	uint32_t hash = simple_hash(id);

	struct cgroup *cg;
	for(cg = cgroup_root; cg ; cg = cg->next) {
		if(hash == cg->hash && strcmp(id, cg->id) == 0)
			break;
	}

	debug(D_CGROUP, "cgroup_find('%s') %s", id, (cg)?"found":"not found");
	return cg;
}

// ----------------------------------------------------------------------------
// detect running cgroups

// callback for find_file_in_subdirs()
void found_subdir_in_dir(const char *dir) {
	debug(D_CGROUP, "examining cgroup dir '%s'", dir);

	struct cgroup *cg = cgroup_find(dir);
	if(!cg) {
		if(*dir && cgroup_max_depth > 0) {
			int depth = 0;
			const char *s;

			for(s = dir; *s ;s++)
				if(unlikely(*s == '/'))
					depth++;

			if(depth > cgroup_max_depth) {
				info("cgroup '%s' is too deep (%d, while max is %d)", dir, depth, cgroup_max_depth);
				return;
			}
		}
		debug(D_CGROUP, "will add dir '%s' as cgroup", dir);
		cg = cgroup_add(dir);
	}

	if(cg) cg->available = 1;
}

void find_dir_in_subdirs(const char *base, const char *this, void (*callback)(const char *)) {
	int enabled = -1;
	if(!this) this = base;
	size_t dirlen = strlen(this), baselen = strlen(base);
	const char *relative_path = &this[baselen];

	DIR *dir = opendir(this);
	if(!dir) return;

	callback(relative_path);

	struct dirent *de = NULL;
	while((de = readdir(dir))) {
		if(de->d_type == DT_DIR
			&& (
				(de->d_name[0] == '.' && de->d_name[1] == '\0')
				|| (de->d_name[0] == '.' && de->d_name[1] == '.' && de->d_name[2] == '\0')
				))
			continue;

		debug(D_CGROUP, "examining '%s/%s'", this, de->d_name);

		if(de->d_type == DT_DIR) {
			if(enabled == -1) {
				const char *r = relative_path;
				if(*r == '\0') r = "/";
				else if (*r == '/') r++;

				// we check for this option here
				// so that the config will not have settings
				// for leaf directories
				char option[FILENAME_MAX + 1];
				snprintfz(option, FILENAME_MAX, "search for cgroups under %s", r);
				option[FILENAME_MAX] = '\0';
				enabled = config_get_boolean("plugin:cgroups", option, 1);
			}

			if(enabled) {
				char *s = malloc(dirlen + strlen(de->d_name) + 2);
				if(s) {
					strcpy(s, this);
					strcat(s, "/");
					strcat(s, de->d_name);
					find_dir_in_subdirs(base, s, callback);
					free(s);
				}
			}
		}
	}

	closedir(dir);
}

void mark_all_cgroups_as_not_available() {
	debug(D_CGROUP, "marking all cgroups as not available");

	struct cgroup *cg;

	// mark all as not available
	for(cg = cgroup_root; cg ; cg = cg->next)
		cg->available = 0;
}

void cleanup_all_cgroups() {
	struct cgroup *cg = cgroup_root, *last = NULL;

	for(; cg ;) {
		if(!cg->available) {

			if(!last)
				cgroup_root = cg->next;
			else
				last->next = cg->next;

			cgroup_free(cg);

			if(!last)
				cg = cgroup_root;
			else
				cg = last->next;
		}
		else {
			last = cg;
			cg = cg->next;
		}
	}
}

void find_all_cgroups() {
	debug(D_CGROUP, "searching for cgroups");

	mark_all_cgroups_as_not_available();

	if(cgroup_enable_cpuacct_stat || cgroup_enable_cpuacct_usage)
		find_dir_in_subdirs(cgroup_cpuacct_base, NULL, found_subdir_in_dir);

	if(cgroup_enable_blkio)
		find_dir_in_subdirs(cgroup_blkio_base, NULL, found_subdir_in_dir);

	if(cgroup_enable_memory)
		find_dir_in_subdirs(cgroup_memory_base, NULL, found_subdir_in_dir);

	// remove any non-existing cgroups
	cleanup_all_cgroups();

	struct cgroup *cg;
	for(cg = cgroup_root; cg ; cg = cg->next) {
		// fprintf(stderr, " >>> CGROUP '%s' (%u - %s) with name '%s'\n", cg->id, cg->hash, cg->available?"available":"stopped", cg->name);

		if(unlikely(!cg->available))
			continue;

		debug(D_CGROUP, "checking paths for cgroup '%s'", cg->id);

		// check for newly added cgroups
		// and update the filenames they read
		char filename[FILENAME_MAX + 1];
		if(cgroup_enable_cpuacct_stat && !cg->cpuacct_stat.filename) {
			snprintfz(filename, FILENAME_MAX, "%s%s/cpuacct.stat", cgroup_cpuacct_base, cg->id);
			cg->cpuacct_stat.filename = strdup(filename);
			debug(D_CGROUP, "cpuacct.stat filename for cgroup '%s': '%s'", cg->id, cg->cpuacct_stat.filename);
		}
		if(cgroup_enable_cpuacct_usage && !cg->cpuacct_usage.filename) {
			snprintfz(filename, FILENAME_MAX, "%s%s/cpuacct.usage_percpu", cgroup_cpuacct_base, cg->id);
			cg->cpuacct_usage.filename = strdup(filename);
			debug(D_CGROUP, "cpuacct.usage_percpu filename for cgroup '%s': '%s'", cg->id, cg->cpuacct_usage.filename);
		}
		if(cgroup_enable_memory && !cg->memory.filename) {
			snprintfz(filename, FILENAME_MAX, "%s%s/memory.stat", cgroup_memory_base, cg->id);
			cg->memory.filename = strdup(filename);
			debug(D_CGROUP, "memory.stat filename for cgroup '%s': '%s'", cg->id, cg->memory.filename);
		}
		if(cgroup_enable_blkio) {
			if(!cg->io_service_bytes.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.io_service_bytes", cgroup_blkio_base, cg->id);
				cg->io_service_bytes.filename = strdup(filename);
				debug(D_CGROUP, "io_service_bytes filename for cgroup '%s': '%s'", cg->id, cg->io_service_bytes.filename);
			}
			if(!cg->io_serviced.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.io_serviced", cgroup_blkio_base, cg->id);
				cg->io_serviced.filename = strdup(filename);
				debug(D_CGROUP, "io_serviced filename for cgroup '%s': '%s'", cg->id, cg->io_serviced.filename);
			}
			if(!cg->throttle_io_service_bytes.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.throttle.io_service_bytes", cgroup_blkio_base, cg->id);
				cg->throttle_io_service_bytes.filename = strdup(filename);
				debug(D_CGROUP, "throttle_io_service_bytes filename for cgroup '%s': '%s'", cg->id, cg->throttle_io_service_bytes.filename);
			}
			if(!cg->throttle_io_serviced.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.throttle.io_serviced", cgroup_blkio_base, cg->id);
				cg->throttle_io_serviced.filename = strdup(filename);
				debug(D_CGROUP, "throttle_io_serviced filename for cgroup '%s': '%s'", cg->id, cg->throttle_io_serviced.filename);
			}
			if(!cg->io_merged.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.io_merged", cgroup_blkio_base, cg->id);
				cg->io_merged.filename = strdup(filename);
				debug(D_CGROUP, "io_merged filename for cgroup '%s': '%s'", cg->id, cg->io_merged.filename);
			}
			if(!cg->io_queued.filename) {
				snprintfz(filename, FILENAME_MAX, "%s%s/blkio.io_queued", cgroup_blkio_base, cg->id);
				cg->io_queued.filename = strdup(filename);
				debug(D_CGROUP, "io_queued filename for cgroup '%s': '%s'", cg->id, cg->io_queued.filename);
			}
		}
	}

	debug(D_CGROUP, "done searching for cgroups");
	return;
}

// ----------------------------------------------------------------------------
// generate charts

#define CHART_TITLE_MAX 300

void update_cgroup_charts(int update_every) {
	debug(D_CGROUP, "updating cgroups charts");

	char type[RRD_ID_LENGTH_MAX + 1];
	char title[CHART_TITLE_MAX + 1];

	struct cgroup *cg;
	RRDSET *st;

	for(cg = cgroup_root; cg ; cg = cg->next) {
		if(!cg->available || !cg->enabled)
			continue;

		if(cg->id[0] == '\0')
			strcpy(type, "cgroup_root");
		else if(cg->id[0] == '/')
			snprintfz(type, RRD_ID_LENGTH_MAX, "cgroup_%s", cg->chart_id);
		else
			snprintfz(type, RRD_ID_LENGTH_MAX, "cgroup_%s", cg->chart_id);

		netdata_fix_chart_id(type);

		if(cg->cpuacct_stat.updated) {
			st = rrdset_find_bytype(type, "cpu");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "CPU Usage for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "cpu", NULL, "cpu", "cgroup.cpu", title, "%", 40000, update_every, RRDSET_TYPE_STACKED);

				rrddim_add(st, "user", NULL, 100, hz, RRDDIM_INCREMENTAL);
				rrddim_add(st, "system", NULL, 100, hz, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "user", cg->cpuacct_stat.user);
			rrddim_set(st, "system", cg->cpuacct_stat.system);
			rrdset_done(st);
		}

		if(cg->cpuacct_usage.updated) {
			char id[RRD_ID_LENGTH_MAX + 1];
			unsigned int i;

			st = rrdset_find_bytype(type, "cpu_per_core");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "CPU Usage Per Core for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "cpu_per_core", NULL, "cpu", "cgroup.cpu_per_core", title, "%", 40100, update_every, RRDSET_TYPE_STACKED);

				for(i = 0; i < cg->cpuacct_usage.cpus ;i++) {
					snprintfz(id, CHART_TITLE_MAX, "cpu%d", i);
					rrddim_add(st, id, NULL, 100, 1000000, RRDDIM_INCREMENTAL);
				}
			}
			else rrdset_next(st);

			for(i = 0; i < cg->cpuacct_usage.cpus ;i++) {
				snprintfz(id, CHART_TITLE_MAX, "cpu%d", i);
				rrddim_set(st, id, cg->cpuacct_usage.cpu_percpu[i]);
			}
			rrdset_done(st);
		}

		if(cg->memory.updated) {
			if(cg->memory.cache + cg->memory.rss + cg->memory.rss_huge + cg->memory.mapped_file > 0) {
				st = rrdset_find_bytype(type, "mem");
				if(!st) {
					snprintfz(title, CHART_TITLE_MAX, "Memory Usage for cgroup %s", cg->chart_title);
					st = rrdset_create(type, "mem", NULL, "mem", "cgroup.mem", title, "MB", 40200, update_every,
					                   RRDSET_TYPE_STACKED);

					rrddim_add(st, "cache", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
					rrddim_add(st, "rss", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
					if(cg->memory.has_dirty_swap)
						rrddim_add(st, "swap", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
					rrddim_add(st, "rss_huge", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
					rrddim_add(st, "mapped_file", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
				}
				else rrdset_next(st);

				rrddim_set(st, "cache", cg->memory.cache);
				rrddim_set(st, "rss", cg->memory.rss);
				if(cg->memory.has_dirty_swap)
					rrddim_set(st, "swap", cg->memory.swap);
				rrddim_set(st, "rss_huge", cg->memory.rss_huge);
				rrddim_set(st, "mapped_file", cg->memory.mapped_file);
				rrdset_done(st);
			}

			st = rrdset_find_bytype(type, "writeback");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Writeback Memory for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "writeback", NULL, "mem", "cgroup.writeback", title, "MB", 40300,
				                   update_every, RRDSET_TYPE_AREA);

				if(cg->memory.has_dirty_swap)
					rrddim_add(st, "dirty", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
				rrddim_add(st, "writeback", NULL, 1, 1024 * 1024, RRDDIM_ABSOLUTE);
			}
			else rrdset_next(st);

			if(cg->memory.has_dirty_swap)
				rrddim_set(st, "dirty", cg->memory.dirty);
			rrddim_set(st, "writeback", cg->memory.writeback);
			rrdset_done(st);

			if(cg->memory.pgpgin + cg->memory.pgpgout > 0) {
				st = rrdset_find_bytype(type, "mem_activity");
				if(!st) {
					snprintfz(title, CHART_TITLE_MAX, "Memory Activity for cgroup %s", cg->chart_title);
					st = rrdset_create(type, "mem_activity", NULL, "mem", "cgroup.mem_activity", title, "MB/s",
					                   40400, update_every, RRDSET_TYPE_LINE);

					rrddim_add(st, "pgpgin", "in", sysconf(_SC_PAGESIZE), 1024 * 1024, RRDDIM_INCREMENTAL);
					rrddim_add(st, "pgpgout", "out", -sysconf(_SC_PAGESIZE), 1024 * 1024, RRDDIM_INCREMENTAL);
				}
				else rrdset_next(st);

				rrddim_set(st, "pgpgin", cg->memory.pgpgin);
				rrddim_set(st, "pgpgout", cg->memory.pgpgout);
				rrdset_done(st);
			}

			if(cg->memory.pgfault + cg->memory.pgmajfault > 0) {
				st = rrdset_find_bytype(type, "pgfaults");
				if(!st) {
					snprintfz(title, CHART_TITLE_MAX, "Memory Page Faults for cgroup %s", cg->chart_title);
					st = rrdset_create(type, "pgfaults", NULL, "mem", "cgroup.pgfaults", title, "MB/s", 40500,
					                   update_every, RRDSET_TYPE_LINE);

					rrddim_add(st, "pgfault", NULL, sysconf(_SC_PAGESIZE), 1024 * 1024, RRDDIM_INCREMENTAL);
					rrddim_add(st, "pgmajfault", "swap", -sysconf(_SC_PAGESIZE), 1024 * 1024, RRDDIM_INCREMENTAL);
				}
				else rrdset_next(st);

				rrddim_set(st, "pgfault", cg->memory.pgfault);
				rrddim_set(st, "pgmajfault", cg->memory.pgmajfault);
				rrdset_done(st);
			}
		}

		if(cg->io_service_bytes.updated && cg->io_service_bytes.Read + cg->io_service_bytes.Write > 0) {
			st = rrdset_find_bytype(type, "io");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "I/O Bandwidth (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "io", NULL, "disk", "cgroup.io", title, "KB/s", 41200,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1024, RRDDIM_INCREMENTAL);
				rrddim_add(st, "write", NULL, -1, 1024, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->io_service_bytes.Read);
			rrddim_set(st, "write", cg->io_service_bytes.Write);
			rrdset_done(st);
		}

		if(cg->io_serviced.updated && cg->io_serviced.Read + cg->io_serviced.Write > 0) {
			st = rrdset_find_bytype(type, "serviced_ops");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Serviced I/O Operations (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "serviced_ops", NULL, "disk", "cgroup.serviced_ops", title, "operations/s", 41200,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1, RRDDIM_INCREMENTAL);
				rrddim_add(st, "write", NULL, -1, 1, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->io_serviced.Read);
			rrddim_set(st, "write", cg->io_serviced.Write);
			rrdset_done(st);
		}

		if(cg->throttle_io_service_bytes.updated && cg->throttle_io_service_bytes.Read + cg->throttle_io_service_bytes.Write > 0) {
			st = rrdset_find_bytype(type, "io");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Throttle I/O Bandwidth (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "io", NULL, "disk", "cgroup.io", title, "KB/s", 41200,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1024, RRDDIM_INCREMENTAL);
				rrddim_add(st, "write", NULL, -1, 1024, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->throttle_io_service_bytes.Read);
			rrddim_set(st, "write", cg->throttle_io_service_bytes.Write);
			rrdset_done(st);
		}


		if(cg->throttle_io_serviced.updated && cg->throttle_io_serviced.Read + cg->throttle_io_serviced.Write > 0) {
			st = rrdset_find_bytype(type, "throttle_serviced_ops");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Throttle Serviced I/O Operations (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "throttle_serviced_ops", NULL, "disk", "cgroup.throttle_serviced_ops", title, "operations/s", 41200,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1, RRDDIM_INCREMENTAL);
				rrddim_add(st, "write", NULL, -1, 1, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->throttle_io_serviced.Read);
			rrddim_set(st, "write", cg->throttle_io_serviced.Write);
			rrdset_done(st);
		}

		if(cg->io_queued.updated) {
			st = rrdset_find_bytype(type, "queued_ops");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Queued I/O Operations (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "queued_ops", NULL, "disk", "cgroup.queued_ops", title, "operations", 42000,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1, RRDDIM_ABSOLUTE);
				rrddim_add(st, "write", NULL, -1, 1, RRDDIM_ABSOLUTE);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->io_queued.Read);
			rrddim_set(st, "write", cg->io_queued.Write);
			rrdset_done(st);
		}

		if(cg->io_merged.updated && cg->io_merged.Read + cg->io_merged.Write > 0) {
			st = rrdset_find_bytype(type, "merged_ops");
			if(!st) {
				snprintfz(title, CHART_TITLE_MAX, "Merged I/O Operations (all disks) for cgroup %s", cg->chart_title);
				st = rrdset_create(type, "merged_ops", NULL, "disk", "cgroup.merged_ops", title, "operations/s", 42100,
				                   update_every, RRDSET_TYPE_LINE);

				rrddim_add(st, "read", NULL, 1, 1024, RRDDIM_INCREMENTAL);
				rrddim_add(st, "write", NULL, -1, 1024, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(st);

			rrddim_set(st, "read", cg->io_merged.Read);
			rrddim_set(st, "write", cg->io_merged.Write);
			rrdset_done(st);
		}
	}

	debug(D_CGROUP, "done updating cgroups charts");
}

// ----------------------------------------------------------------------------
// cgroups main

int do_sys_fs_cgroup(int update_every, unsigned long long dt) {
	static int cgroup_global_config_read = 0;
	static time_t last_run = 0;
	time_t now = time(NULL);

	if(dt) {};

	if(unlikely(!cgroup_global_config_read)) {
		read_cgroup_plugin_configuration();
		cgroup_global_config_read = 1;
	}

	if(unlikely(cgroup_enable_new_cgroups_detected_at_runtime && now - last_run > cgroup_check_for_new_every)) {
		find_all_cgroups();
		last_run = now;
	}

	read_all_cgroups(cgroup_root);
	update_cgroup_charts(update_every);

	return 0;
}

void *cgroups_main(void *ptr)
{
	if(ptr) { ; }

	info("CGROUP Plugin thread created with task id %d", gettid());

	if(pthread_setcanceltype(PTHREAD_CANCEL_DEFERRED, NULL) != 0)
		error("Cannot set pthread cancel type to DEFERRED.");

	if(pthread_setcancelstate(PTHREAD_CANCEL_ENABLE, NULL) != 0)
		error("Cannot set pthread cancel state to ENABLE.");

	struct rusage thread;

	// when ZERO, attempt to do it
	int vdo_sys_fs_cgroup 			= 0;
	int vdo_cpu_netdata 			= !config_get_boolean("plugin:cgroups", "cgroups plugin resources", 1);

	// keep track of the time each module was called
	unsigned long long sutime_sys_fs_cgroup = 0ULL;

	// the next time we will run - aligned properly
	unsigned long long sunext = (time(NULL) - (time(NULL) % rrd_update_every) + rrd_update_every) * 1000000ULL;
	unsigned long long sunow;

	RRDSET *stcpu_thread = NULL;

	for(;1;) {
		if(unlikely(netdata_exit)) break;

		// delay until it is our time to run
		while((sunow = timems()) < sunext)
			usleep((useconds_t)(sunext - sunow));

		// find the next time we need to run
		while(timems() > sunext)
			sunext += rrd_update_every * 1000000ULL;

		if(unlikely(netdata_exit)) break;

		// BEGIN -- the job to be done

		if(!vdo_sys_fs_cgroup) {
			debug(D_PROCNETDEV_LOOP, "PROCNETDEV: calling do_sys_fs_cgroup().");
			sunow = timems();
			vdo_sys_fs_cgroup = do_sys_fs_cgroup(rrd_update_every, (sutime_sys_fs_cgroup > 0)?sunow - sutime_sys_fs_cgroup:0ULL);
			sutime_sys_fs_cgroup = sunow;
		}
		if(unlikely(netdata_exit)) break;

		// END -- the job is done

		// --------------------------------------------------------------------

		if(!vdo_cpu_netdata) {
			getrusage(RUSAGE_THREAD, &thread);

			if(!stcpu_thread) stcpu_thread = rrdset_find("netdata.plugin_cgroups_cpu");
			if(!stcpu_thread) {
				stcpu_thread = rrdset_create("netdata", "plugin_cgroups_cpu", NULL, "proc.internal", NULL, "NetData CGroups Plugin CPU usage", "milliseconds/s", 132000, rrd_update_every, RRDSET_TYPE_STACKED);

				rrddim_add(stcpu_thread, "user",  NULL,  1, 1000, RRDDIM_INCREMENTAL);
				rrddim_add(stcpu_thread, "system", NULL, 1, 1000, RRDDIM_INCREMENTAL);
			}
			else rrdset_next(stcpu_thread);

			rrddim_set(stcpu_thread, "user"  , thread.ru_utime.tv_sec * 1000000ULL + thread.ru_utime.tv_usec);
			rrddim_set(stcpu_thread, "system", thread.ru_stime.tv_sec * 1000000ULL + thread.ru_stime.tv_usec);
			rrdset_done(stcpu_thread);
		}
	}

	pthread_exit(NULL);
	return NULL;
}