linux threads and services

	A	B	C	D	E	F	G	H	I
1	uob	RHEL5	RHEL6	RHEL7	name	short description	used for	RPM	long description, external links

3	K	y			aio	handles asynchronous IO		kernel
5	K		y		ata_aux			kernel
6	K	y	y		ata/2			kernel
9	K	y			beiscsi_q_irq4			kernel
10	K			y	bioset	Used for block I/O - please refer to https://lwn.net/Articles/26404/
11	K	y			bond0	handles link aggregation (bonding) of bond0 interface	bonding	kernel
12	K	y			bond1	handles link aggregation (bonding) of bond1 interface	bonding	kernel
13	K	n	y		cgroup
14	K	y			cnic_wq	QLogic Linux qlcnic NIC Driver (bnx2) - write queue	hw driver NIC	kernel
15	K	y			cqueue			kernel
17	K			y	deferwq	??
18	K	y			delete_workqueue			kernel
19	K	y			dlm_astd	distributed lock manager - used in Red Hat Clustering	RHCS	cman
20	K	y			dlm_controld	distributed lock manager - used in Red Hat Clustering	RHCS	cman
21	K	y			dlm_recv	distributed lock manager - used in Red Hat Clustering	RHCS	cman
22	K	y			dlm_scand	distributed lock manager - used in Red Hat Clustering	RHCS	cman
23	K	y			dlm_send	distributed lock manager - used in Red Hat Clustering	RHCS	cman
25	K	y			fc_dl		Fiber Channel	kernel
26	K	y			fc_wq		Fiber Channel	kernel
27	K	y			gfs2_logd	GFS2 journalling?	GFS2	kernel
28	K	y			gfs2_quotad	GFS2 quota control thread	GFS2	kernel
29	K	y			glock_workqueue	GFS2 glock locking	GFS2	kernel	/Documentation/filesystems/gfs2-glocks.txt
37	K	y			ib_addr	Infiniband address resolution?	Infiniband	kernel
38	K	y			ib_cb		Infiniband	kernel
39	K	y			ib_inform		Infiniband	kernel
40	K	y			ib_multicast		Infiniband	kernel
46	K		y		jbd2/dm-0-8	journal block device for device dm-0
47	K				jbd2/sda1-8	journal block device for device sda1
48	K	y	y		kacpid	deals with ACPI, e.g. power saving functions	ACPI	kernel
49	K	y			kauditd	kernel thread for handling audit events		kernel
50	K	y			kblockd	used for running low-level disk operations	disk	kernel
51	K				kdmflush	is used by Device Mapper to process deferred work	DM
53	K	y			kedac	kstriped		kernel	notice that this is not only ECC-memory checks. Some architectures have ECC detectors for L1, L2 and L3 caches, along with DMA engines, fabric switches, main data path switches, interconnections, and various other hardware data paths. If the hardware reports it, then a edac_device device probably can be constructed to harvest and present that to userspace. /Documentation/edac.txt
54	K	y			khelper	used as a scheduleable context for stuff that call_user_mode_helper wants to run - like /sbin/hotplug or modprobe		kernel
55	K	y			khubd	used for configuring USB	USB	kernel
56	K		y		khugepaged	transparent hugepage support			/sys/kernel/mm/transparent_hugepage
57	K	y			khungtaskd	hung task detect		kernel	The Detect Hung Task kernel thread (khungtaskd) is present in all RHEL since RHEL5.5 up (>2.6.18-194), providing the ability to detect tasks stuck in D-State longer than a specified time period (set in /proc/sys/kernel/hung_task_timeout_sec, 120 seconds by default). https://access.redhat.com/knowledge/solutions/31453
59	K		y		kintegrity/1			kernel
60	K	y			kipmi0			kernel
61	K	y			kjournald	helper thread for handling filesystem journals		kernel
62	K	y			kmpathd	helper thread for multipath daemon	multipath disk	kernel
63	K	y	y		kondemand	helper thread for ondemand cpu frequency regulator	cpu frequency govenor	kernel
64	K	y			kpsmoused	handles mouse IO		kernel
66	K	y			kseriod	used for handling serial io		kernel
67	K		y		kslowd000
68	K		y		kslowd001
69	K	y			ksnapd	helper thread for handling dm-snaps	DM	kernel
70	K	y			ksoftirqd/0	kernel healper thread to handle softirqs that can't be handled immediately - running on CPU0		kernel
71	K	y			ksoftirqd/1	kernel healper thread to handle softirqs that can't be handled immediately - running on CPU1		kernel
72	K	y			kstriped	helper thread for striped volumes		kernel
73	K		y		ksuspend_usbd
74	K	y			kswapd	makes sure there is enough free memory	paging and swapping	kernel	when memory is exhausted, kswapd reclaims pagecache pages and/or swaps in.
78	K	y	y		kworker	delayed work not handed by another thread, used throughout kernel		kernel
79	K	y			local_sa		Infiniband	kernel
80	K	y			lock_dlm	distributed lock manager - used in Red Hat Clustering	DLM	kernel
87	K	y			migration	moves processes between CPUs - for load balancing		kernel
104	K	y			pdflush	pdflush does the laundry of dirty pages. (sends dirty memory to disk)	paging	kernel	Earlier versions of Linux used a kernel thread called bdflush to systematically scan the page cache looking for dirty pages to flush, and they used a second kernel thread called kupdate to ensure that no page remains dirty for too long. Linux 2.6 has replaced both of them with a group of general purpose kernel threads called pdflush. There must be at least two pdflush kernel threads and at most eight. pdflush thread is created every dirty_writeback_centisecs to write dirty pages older than dirty_expire_centisecs to disk. pdflush will be woken up earlier than dirty_writeback_centisecs if dirty_background_ratio is exceeded. When woken, pdflush will start flushing pages in the background. if dirty_ratio is reached, then then no new pages will be cached. All system processes will start operating in the synchronious mode of writing, leading to lots of D states (I/O-wait). All percentages are of physical memory (MemTotal in /proc/meminfo) pdflush, short for page dirty flush, was replaced by the flusher threads in Linux kernel version 2.6.32. pdflush dynamically manages its gang of threads based on the amount of I/O pending. More work, more threads. Makes sense, right? The problem is that threads represent units of processing power, but disk I/O is bottlenecked by disk throughput. A bunch of threads hammering a single disk doesn't make sense, particularly if some other disk is left idle. This design flaw is remedied by the flusher threads, which spawn one thread per disk spindle, and are designed to maximize throughput.
105	K	y			ping
113	K	y			qla2xxx_0_dpc	qla device driver thread		kernel
116	K	y			rdma_cm			kernel
123	K	y			rpciod/3			kernel
127	K	y			scsi_eh	SCSI event handler	SCSI	kernel
128	K	y			scsi_wq	SCSI work queue handler	SCSI	kernel
135	K		y		sync_supers			kernel
138	K	y			watchdog	notices if the systems appears to be hung events		kernel
144	K	y
145	K	y
146	K	y
147	K	y
148	K	y
149	K	y
150	K	y
151	K	y
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177