1ipmipower(8) System Commands ipmipower(8)
2
6 ipmipower - IPMI power control utility
7
9 ipmipower [OPTION...]
10
12 Ipmipower allows users to remotely power on, off, cycle, hard reset,
13 get a power status query, perform a pulse diagnostic interrupt, or ini‐
14 tiate a soft-shutdown of the OS via ACPI through the IPMI over LAN pro‐
15 tocol.
16 When a power command (--on, --off, --cycle, --reset, --stat, --pulse,
18 or --soft) is specified on the command line, Ipmipower will attempt to
19 run the power command on all hostnames listed on the command line then
20 exit.
21 If no power commands are specified on the command line, ipmipower will
23 run in interactive mode. Interactive mode gives the user a command line
24 interface to enter various commands. Details of the interactive command
25 line interface can be found below under INTERACTIVE COMMANDS.
26 Listed below are general IPMI options, tool specific options, trouble
28 shooting information, workaround information, examples, and known is‐
29 sues. For a general introduction to FreeIPMI please see freeipmi(7).
30
32 The following options are general options for configuring IPMI communi‐
33 cation and executing general tool commands.
34 -D IPMIDRIVER, --driver-type=IPMIDRIVER
36 Specify the driver type to use instead of doing an auto selec‐
37 tion. The currently available outofband drivers are LAN and
38 LAN_2_0, which perform IPMI 1.5 and IPMI 2.0 respectively.
39 -h IPMIHOST1,IPMIHOST2,..., --hostname=IPMIHOST1[:PORT],IPMI‐
41 HOST2[:PORT],...
42 Specify the remote host(s) to communicate with. Multiple host‐
43 names may be separated by comma or may be specified in a range
44 format; see HOSTRANGED SUPPORT below. An optional port can be
45 specified with each host, which may be useful in port forwarding
46 or similar situations. If specifying an IPv6 address and port,
47 use the format [ADDRESS]:PORT.
48 -u USERNAME, --username=USERNAME
50 Specify the username to use when authenticating with the remote
51 host. If not specified, a null (i.e. anonymous) username is as‐
52 sumed. The user must have atleast OPERATOR privileges to run the
53 --on, --off, --reset, --cycle, --pulse, or --soft power control
54 commands. The user must have atleast USER privileges to deter‐
55 mine the power status of the machine through --stat.
56 -p PASSWORD, --password=PASSWORD
58 Specify the password to use when authenticationg with the remote
59 host. If not specified, a null password is assumed. Maximum
60 password length is 16 for IPMI 1.5 and 20 for IPMI 2.0.
61 -P, --password-prompt
63 Prompt for password to avoid possibility of listing it in
64 process lists.
65 -k K_G, --k-g=K_G
67 Specify the K_g BMC key to use when authenticating with the re‐
68 mote host for IPMI 2.0. If not specified, a null key is assumed.
69 To input the key in hexadecimal form, prefix the string with
70 '0x'. E.g., the key 'abc' can be entered with the either the
71 string 'abc' or the string '0x616263'
72 -K, --k-g-prompt
74 Prompt for k-g to avoid possibility of listing it in process
75 lists.
76 --session-timeout=MILLISECONDS
78 Specify the session timeout in milliseconds. Defaults to 20000
79 milliseconds (20 seconds) if not specified.
80 --retransmission-timeout=MILLISECONDS
82 Specify the packet retransmission timeout in milliseconds. De‐
83 faults to 400 milliseconds (0.4 seconds) if not specified.
84 -a AUTHENTICATION-TYPE, --authentication-type=AUTHENTICATION-TYPE
86 Specify the IPMI 1.5 authentication type to use. The currently
87 available authentication types are NONE, STRAIGHT_PASSWORD_KEY,
88 MD2, and MD5. Defaults to MD5 if not specified.
89 -I CIPHER-SUITE-ID, --cipher-suite-id=CIPHER-SUITE-ID
91 Specify the IPMI 2.0 cipher suite ID to use. The Cipher Suite ID
92 identifies a set of authentication, integrity, and confidential‐
93 ity algorithms to use for IPMI 2.0 communication. The authenti‐
94 cation algorithm identifies the algorithm to use for session
95 setup, the integrity algorithm identifies the algorithm to use
96 for session packet signatures, and the confidentiality algorithm
97 identifies the algorithm to use for payload encryption. Defaults
98 to cipher suite ID 3 if not specified. The following cipher
99 suite ids are currently supported:
100 0 - Authentication Algorithm = None; Integrity Algorithm = None;
102 Confidentiality Algorithm = None
103 1 - Authentication Algorithm = HMAC-SHA1; Integrity Algorithm =
105 None; Confidentiality Algorithm = None
106 2 - Authentication Algorithm = HMAC-SHA1; Integrity Algorithm =
108 HMAC-SHA1-96; Confidentiality Algorithm = None
109 3 - Authentication Algorithm = HMAC-SHA1; Integrity Algorithm =
111 HMAC-SHA1-96; Confidentiality Algorithm = AES-CBC-128
112 6 - Authentication Algorithm = HMAC-MD5; Integrity Algorithm =
114 None; Confidentiality Algorithm = None
115 7 - Authentication Algorithm = HMAC-MD5; Integrity Algorithm =
117 HMAC-MD5-128; Confidentiality Algorithm = None
118 8 - Authentication Algorithm = HMAC-MD5; Integrity Algorithm =
120 HMAC-MD5-128; Confidentiality Algorithm = AES-CBC-128
121 11 - Authentication Algorithm = HMAC-MD5; Integrity Algorithm =
123 MD5-128; Confidentiality Algorithm = None
124 12 - Authentication Algorithm = HMAC-MD5; Integrity Algorithm =
126 MD5-128; Confidentiality Algorithm = AES-CBC-128
127 15 - Authentication Algorithm = HMAC-SHA256; Integrity Algorithm
129 = None; Confidentiality Algorithm = None
130 16 - Authentication Algorithm = HMAC-SHA256; Integrity Algorithm
132 = HMAC_SHA256_128; Confidentiality Algorithm = None
133 17 - Authentication Algorithm = HMAC-SHA256; Integrity Algorithm
135 = HMAC_SHA256_128; Confidentiality Algorithm = AES-CBC-128
136 -l PRIVILEGE-LEVEL, --privilege-level=PRIVILEGE-LEVEL
138 Specify the privilege level to be used. The currently available
139 privilege levels are USER, OPERATOR, and ADMIN. Defaults to OP‐
140 ERATOR if not specified.
141 --config-file=FILE
143 Specify an alternate configuration file.
144 -W WORKAROUNDS, --workaround-flags=WORKAROUNDS
146 Specify workarounds to vendor compliance issues. Multiple work‐
147 arounds can be specified separated by commas. A special command
148 line flag of "none", will indicate no workarounds (may be useful
149 for overriding configured defaults). See WORKAROUNDS below for a
150 list of available workarounds.
151 --debug
153 Turn on debugging.
154 -?, --help
156 Output a help list and exit.
157 --usage
159 Output a usage message and exit.
160 -V, --version
162 Output the program version and exit.
163
165 The following options are specific to ipmipower.
166 -n, --on
168 Power on the target hosts.
169 -f, --off
171 Power off the target hosts.
172 -c, --cycle
174 Power cycle the target hosts.
175 -r, --reset
177 Reset the target hosts.
178 -s, --stat
180 Get power status of the target hosts.
181 --pulse
183 Send power diagnostic interrupt to target hosts.
184 --soft Initiate a soft-shutdown of the OS via ACPI.
186 --on-if-off
188 The IPMI specification does not require the power cycle or hard
189 reset commands to turn on a machine that is currently powered
190 off. This option will force ipmipower to issue a power on com‐
191 mand instead of a power cycle or hard reset command if the re‐
192 mote machine's power is currently off.
193 --wait-until-on
195 The IPMI specification allows power on commands to return prior
196 to the power on actually taking place. This option will force
197 ipmipower to regularly query the remote BMC and return only af‐
198 ter the machine has powered on.
199 --wait-until-off
201 The IPMI specification allows power off commands to return prior
202 the power off actually taking place. This option will force ip‐
203 mipower to regularly query the remote BMC and return only after
204 the machine has powered off.
205 --oem-power-type=OEM-POWER-TYPE
207 This option informs ipmipower to initiate power control opera‐
208 tions via an IPMI OEM specific power control extension. The cur‐
209 rently available POWERTYPEs are NONE and C410X. Please see OEM
210 POWER EXTENSIONS below for additional information.
211
213 The following options are used to change the networking behavior of ip‐
214 mipower.
215 --retransmission-wait-timeout=MILLISECONDS
217 Specify the retransmission wait timeout length in milliseconds.
218 The retransmission wait timeout is similar to the retransmission
219 timeout above, but is used specifically for power completion
220 verification with the --wait-until-on and --wait-until-off op‐
221 tions. Defaults to 500 milliseconds (0.5 seconds).
222 --retransmission-backoff-count=COUNT
224 Specify the retransmission backoff count for retransmissions.
225 After ever COUNT retransmissions, the retransmission timeout
226 length will be increased by another factor. Defaults to 8.
227 --ping-interval=MILLISECONDS
229 Specify the ping interval length in milliseconds. When running
230 in interactive mode, RMCP (Remote Management Control Protocol)
231 discovery messages will be sent to all configured remote hosts
232 every MILLISECONDS to confirm their support of IPMI. Power com‐
233 mands cannot be sent to a host until it is discovered (or re-
234 discovered if previously lost). Defaults to 5000 milliseconds (5
235 seconds). Ping discovery messages can be disabled by setting
236 this valu to 0. RMCP ping discovery messages are automatically
237 disabled in non-interactive mode.
238 --ping-timeout=MILLISECONDS
240 Specify the ping timeout length in milliseconds. When running in
241 interactive mode, RMCP (Remote Management Control Protocol) mes‐
242 sages discovery will be sent to all configured remote hosts to
243 confirm their support of IPMI. A remote host is considered
244 undiscovered if the host does not respond in MILLISECONDS time.
245 Defaults to 30000 milliseconds (30 seconds). The ping timeout
246 cannot be larger than the ping interval.
247 --ping-packet-count=COUNT
249 Specify the ping packet count size. Defaults to 10. See the
250 --ping-percent-fR option below for more information on this op‐
251 tion.
252 --ping-percent=PERCENT
254 Specify the ping percent value. Defaults to 50. Since IPMI is
255 based on UDP, it is difficult for ipmipower to distinguish be‐
256 tween a missing machine and a bad (or heavily loaded) network
257 connection in interactive mode. when running in interactive
258 mode. For example, suppose a link consistently drops 80% of the
259 packets to a particular machine. The power control operation may
260 have difficulty completing, although a recent pong response from
261 RMCP makes ipmipower believe the machine is up and functioning
262 properly. The ping packet count and percent options are used to
263 alleviate this problem. Ipmipower will monitor RMCP ping pack‐
264 ets in packet count chunks. If ipmipower does not receive a re‐
265 sponse to greater than ping percent of those packets, ipmipower
266 will assume the link to this node is bad and will not send power
267 control operations to that node until the connection is deter‐
268 mined to be reliable. This heuristic can be disabled by setting
269 either the ping packet count or ping percent to 0. This feature
270 is not used if ping interval is set to 0.
271 --ping-consec-count=COUNT
273 Specify the ping consecutive count. This is another heuristic
274 used to determine if a node should be considered discovered,
275 undiscovered, or with a bad connection. If a valid RMCP pong re‐
276 sponse was received for the last COUNT ping packets, a node will
277 be considered discovered, regardless of other heuristics listed
278 above. Defaults to 5. This heuristic can be disabled by setting
279 this value to 0. This feature is not used if other ping features
280 described above are disabled.
281
283 The following options manipulate hostranged output. See HOSTRANGED SUP‐
284 PORT below for additional information on hostranges.
285 -B, --buffer-output
287 Buffer hostranged output. For each node, buffer standard output
288 until the node has completed its IPMI operation. When specifying
289 this option, data may appear to output slower to the user since
290 the the entire IPMI operation must complete before any data can
291 be output. See HOSTRANGED SUPPORT below for additional informa‐
292 tion.
293 -C, --consolidate-output
295 Consolidate hostranged output. The complete standard output from
296 every node specified will be consolidated so that nodes with
297 identical output are not output twice. A header will list those
298 nodes with the consolidated output. When this option is speci‐
299 fied, no output can be seen until the IPMI operations to all
300 nodes has completed. If the user breaks out of the program
301 early, all currently consolidated output will be dumped. See
302 HOSTRANGED SUPPORT below for additional information.
303 -F NUM, --fanout=NUM
305 Specify multiple host fanout. Indicates the maximum number of
306 power control operations that can be executed in parallel.
307 -E, --eliminate
309 Eliminate hosts determined as undetected by ipmidetect. This
310 attempts to remove the common issue of hostranged execution tim‐
311 ing out due to several nodes being removed from service in a
312 large cluster. The ipmidetectd daemon must be running on the
313 node executing the command.
314 --always-prefix
316 Always prefix output, even if only one host is specified or com‐
317 municating in-band. This option is primarily useful for script‐
318 ing purposes. Option will be ignored if specified with the -C
319 option.
320
322 Ipmipower provides the following interactive commands at the ipmipower>
323 prompt. Before any power commands (on, off, cycle, reset, stat, pulse,
324 or soft) can be used, hostnames must be configured into ipmipower, ei‐
325 ther through the command prompt or the hostname command below. The pa‐
326 rameters and options to the commands below mirror their appropriate
327 command line options.
328 hostname [IPMIHOST(s)]
330 Specify a new set of hosts. No input to unconfigure all hosts.
331 username [USERNAME]
333 Specify a new username. No input for null username.
334 password [PASSWORD]
336 Specify a new password. No input for null password.
337 k_g [K_G]
339 Specify a new K_g BMC Key. No input for null key. Prefix with
340 '0x' to enter a key in hexadecimal
341 ipmi-version IPMIVERSION
343 Specify the ipmi version to use.
344 session-timeout MILLISECONDS
346 Specify a new session timeout length.
347 retransmission-timeout MILLISECONDS
349 Specify a new retransmiision timeout length.
350 authentication-type AUTHENTICATION-TYPE
352 Specify the authentication type to use.
353 cipher-suite-id CIPHER-SUITE-ID
355 Specify the cipher suite id to use.
356 privilege-level PRIVILEGE-LEVEL
358 Specify the privilege level to use.
359 workaround-flags WORKAROUNDS
361 Specify workaround flags.
362 debug [on|off]
364 Toggle debug output.
365 on [IPMIHOST(s)]
367 Turn on all configured hosts or specified hosts.
368 off [IPMIHOST(s)]
370 Turn off all configured hosts or specified hosts.
371 cycle [IPMIHOST(s)]
373 Power cycle all configured hosts or specified hosts.
374 reset [IPMIHOST(s)]
376 Reset all configured hosts or specified hosts.
377 stat [IPMIHOST(s)]
379 Query power status for all configured hosts or specified hosts.
380 pulse [IPMIHOST(s)]
382 Pulse diagnostic interrupt all configured hosts or specified
383 hosts.
384 soft [IPMIHOST(s)]
386 Initiate a soft-shutdown for all configured hosts or specified
387 hosts.
388 identify-on [IPMIHOST(s)]
390 Turn on physical system identification.
391 identify-off [IPMIHOST(s)]
393 Turn off physical system identification.
394 identify-status [IPMIHOST(s)]
396 Query physical system identification status.
397 on-if-off [on|off]
399 Toggle on-if-off functionality.
400 wait-until-on [on|off]
402 Toggle wait-until-on functionality.
403 wait-until-off [on|off]
405 Toggle wait-until-off functionality.
406 retransmission-wait-timeout MILLISECONDS
408 Specify a new retransmission wait timeout length.
409 retransmission-backoff-count COUNT
411 Specify a new retransmission backoff count.
412 ping-interval MILLISECONDS
414 Specify a new ping interval length.
415 ping-timeout MILLISECONDS
417 Specify a new ping timeout length.
418 ping-packet-count COUNT
420 Specify a new ping packet count.
421 ping-percent PERCENT
423 Specify a new ping percent.
424 ping-consec-count COUNT
426 Specify a new ping consec count.
427 buffer-output [on|off]
429 Toggle buffer-output functionality.
430 consolidate-output [on|off]
432 Toggle consolidate-output functionality.
433 fanout COUNT
435 Specify a fanout.
436 always-prefix [on|off]
438 Toggle always-prefix functionality.
439 help Output help menu.
441 version
443 Output version.
444 config Output the current configuration.
446 quit Quit program.
448
450 Some motherboards include IPMI OEM extensions for alternate power con‐
451 trol mechanisms. For example, these power control mechanisms may allow
452 you to power control a sub-device within the system rather than the en‐
453 tire system itself.
454 By specifying an OEM power type via --oem-power-type on the command
456 line or freeipmi.conf(5), you can instruct ipmipower to execute alter‐
457 nate power control implementations over the standard ones. Depending on
458 the OEM extension, some power control commands may no longer be avail‐
459 able. For example, an OEM extension may allow on but not cycle. Spe‐
460 cific ipmipower options may not longer function either.
461 Some OEM extensions may require additional arguments for their power
463 control action, such as a sub-device identifier. Additional arguments
464 can be provided by appending a plus sign ('+') and the extra informa‐
465 tion to the end of the hostname. This can be done on the command line
466 or in interactive mode. For example, the hostname mynode+18 would indi‐
467 cate the power control operation should be sent to the host mynode, and
468 18 is the identifier of a possible sub-device to be power controlled.
469 The --consolidate-output option is commonly disabled when using an OEM
470 power control that requires extra arguments.
471 Because OEM power control may involve subtypes, it is possible a user
473 may wish to power control multiple sub-devices on the same host. For
474 example, you might specify the hosts mynode+1,mynode+2, indicating to
475 power control subdevice 1 and 2 on mynode. Because many BMCs cannot
476 handle multiple IPMI sessions, power control operations to the same
477 host will be serialized internally by ipmipower.
478 The following are the current OEM power types available, along with in‐
480 formation on the systems they work with and the power control opera‐
481 tions available.
482 C410X This OEM power type supports the power control of PCIe slots on
484 Dell Poweredge C410x systems. It supports on, off, and stat. The
485 PCIe slot number ranges from 1-16 and must always be specified
486 when attempting to power control with this extension. For exam‐
487 ple, the hostname mynode+2 would inform ipmipower to operate on
488 slot number 2 on mynode. The C410x appears to have difficulty
489 handling new slot power control requests until prior ones have
490 completed. Users may wish to strongly consider using the
491 --wait-until-on and --wait-until-off options if multiple slots
492 will be power controlled in short succession.
493 NONE This informs ipmipower that no OEM power type extension is to be
495 used and standard IPMI power control is used. This is the de‐
496 fault.
497
499 Multiple hosts can be input either as an explicit comma separated lists
500 of hosts or a range of hostnames in the general form: prefix[n-m,l-
501 k,...], where n < m and l < k, etc. The later form should not be con‐
502 fused with regular expression character classes (also denoted by []).
503 For example, foo[19] does not represent foo1 or foo9, but rather repre‐
504 sents a degenerate range: foo19.
505 This range syntax is meant only as a convenience on clusters with a
507 prefixNN naming convention and specification of ranges should not be
508 considered necessary -- the list foo1,foo9 could be specified as such,
509 or by the range foo[1,9].
510 Some examples of range usage follow:
512 foo[01-05] instead of foo01,foo02,foo03,foo04,foo05
513 foo[7,9-10] instead of foo7,foo9,foo10
514 foo[0-3] instead of foo0,foo1,foo2,foo3
515 As a reminder to the reader, some shells will interpret brackets ([ and
517 ]) for pattern matching. Depending on your shell, it may be necessary
518 to enclose ranged lists within quotes.
519 When multiple hosts are specified by the user, a socket will be created
521 for each host and polled on, effectively allowing communication to all
522 hosts in parallel. This will allow communication to large numbers of
523 nodes far more quickly than if done in serial. The -F option can con‐
524 figure the number of nodes that can be communicated with in parallel at
525 the same time.
526 By default, standard output from each node specified will be output
528 with the hostname prepended to each line. Although this output is read‐
529 able in many situations, it may be difficult to read in other situa‐
530 tions. For example, output from multiple nodes may be mixed together.
531 The -B and -C options can be used to change this default.
532
534 Determine the power status of foo[0-2] with null username and password
535 ipmipower -h foo[0-2] --stat
536 Determine the power status of foo[0-2] with non-null username and pass‐
538 word
539 ipmipower -h foo[0-2] -u foo -p bar --stat
540 Hard reset nodes foo[0-2] with non-null username and password
542 ipmipower -h foo[0-2] -u foo -p bar --reset
543
545 Most often, IPMI problems are due to configuration problems.
546 IPMI over LAN problems involve a misconfiguration of the remote ma‐
548 chine's BMC. Double check to make sure the following are configured
549 properly in the remote machine's BMC: IP address, MAC address, subnet
550 mask, username, user enablement, user privilege, password, LAN privi‐
551 lege, LAN enablement, and allowed authentication type(s). For IPMI 2.0
552 connections, double check to make sure the cipher suite privilege(s)
553 and K_g key are configured properly. The ipmi-config(8) tool can be
554 used to check and/or change these configuration settings.
555 In addition to the troubleshooting tips below, please see WORKAROUNDS
557 below to also if there are any vendor specific bugs that have been dis‐
558 covered and worked around.
559 Listed below are many of the common issues for error messages. For ad‐
561 ditional support, please e-mail the <freeipmi-users@gnu.org> mailing
562 list.
563 "username invalid" - The username entered (or a NULL username if none
565 was entered) is not available on the remote machine. It may also be
566 possible the remote BMC's username configuration is incorrect.
567 "password invalid" - The password entered (or a NULL password if none
569 was entered) is not correct. It may also be possible the password for
570 the user is not correctly configured on the remote BMC.
571 "password verification timeout" - Password verification has timed out.
573 A "password invalid" error (described above) or a generic "session
574 timeout" (described below) occurred. During this point in the protocol
575 it cannot be differentiated which occurred.
576 "k_g invalid" - The K_g key entered (or a NULL K_g key if none was en‐
578 tered) is not correct. It may also be possible the K_g key is not cor‐
579 rectly configured on the remote BMC.
580 "privilege level insufficient" - An IPMI command requires a higher user
582 privilege than the one authenticated with. Please try to authenticate
583 with a higher privilege. This may require authenticating to a different
584 user which has a higher maximum privilege.
585 "privilege level cannot be obtained for this user" - The privilege
587 level you are attempting to authenticate with is higher than the maxi‐
588 mum allowed for this user. Please try again with a lower privilege. It
589 may also be possible the maximum privilege level allowed for a user is
590 not configured properly on the remote BMC.
591 "authentication type unavailable for attempted privilege level" - The
593 authentication type you wish to authenticate with is not available for
594 this privilege level. Please try again with an alternate authentication
595 type or alternate privilege level. It may also be possible the avail‐
596 able authentication types you can authenticate with are not correctly
597 configured on the remote BMC.
598 "cipher suite id unavailable" - The cipher suite id you wish to authen‐
600 ticate with is not available on the remote BMC. Please try again with
601 an alternate cipher suite id. It may also be possible the available ci‐
602 pher suite ids are not correctly configured on the remote BMC.
603 "ipmi 2.0 unavailable" - IPMI 2.0 was not discovered on the remote ma‐
605 chine. Please try to use IPMI 1.5 instead.
606 "connection timeout" - Initial IPMI communication failed. A number of
608 potential errors are possible, including an invalid hostname specified,
609 an IPMI IP address cannot be resolved, IPMI is not enabled on the re‐
610 mote server, the network connection is bad, etc. Please verify configu‐
611 ration and connectivity.
612 "session timeout" - The IPMI session has timed out. Please reconnect.
614 If this error occurs often, you may wish to increase the retransmission
615 timeout. Some remote BMCs are considerably slower than others.
616 "internal IPMI error" - An IPMI error has occurred that FreeIPMI does
618 not know how to handle. Please e-mail <freeipmi-users@gnu.org> to re‐
619 port the issue.
620
622 When powering on a powered off machine, the client must have a means by
623 which to resolve the MAC address of the remote machine's ethernet card.
624 While most modern IPMI solutions support the ability to ARP and resolve
625 addresses when the machine is powered off, some older machines do not.
626 This is typically solved in one of two ways:
627 1) Enable gratuitous ARPs on the remote machine. The remote machine
629 will send out a gratuitous ARP, which advertises the ethernet IP and
630 MAC address so that other machines on the network this information
631 their local ARP cache. For large clusters, this method is not recom‐
632 mended since gratuitous ARPs can flood the network with unnecessary
633 traffic.
634 2) Permanently store the remote machine's MAC address in the local ARP
636 cache. This is the more common approach on large clusters.
637 Other methods are listed in the IPMI specification.
639
641 With so many different vendors implementing their own IPMI solutions,
642 different vendors may implement their IPMI protocols incorrectly. The
643 following describes a number of workarounds currently available to han‐
644 dle discovered compliance issues. When possible, workarounds have been
645 implemented so they will be transparent to the user. However, some will
646 require the user to specify a workaround be used via the -W option.
647 The hardware listed below may only indicate the hardware that a problem
649 was discovered on. Newer versions of hardware may fix the problems in‐
650 dicated below. Similar machines from vendors may or may not exhibit the
651 same problems. Different vendors may license their firmware from the
652 same IPMI firmware developer, so it may be worthwhile to try work‐
653 arounds listed below even if your motherboard is not listed.
654 If you believe your hardware has an additional compliance issue that
656 needs a workaround to be implemented, please contact the FreeIPMI main‐
657 tainers on <freeipmi-users@gnu.org> or <freeipmi-devel@gnu.org>.
658 authcap - This workaround flag will skip early checks for username ca‐
660 pabilities, authentication capabilities, and K_g support and allow IPMI
661 authentication to succeed. It works around multiple issues in which the
662 remote system does not properly report username capabilities, authenti‐
663 cation capabilities, or K_g status. Those hitting this issue may see
664 "username invalid", "authentication type unavailable for attempted
665 privilege level", or "k_g invalid" errors. Issue observed on Asus
666 P5M2/P5MT-R/RS162-E4/RX4, Intel SR1520ML/X38ML, and Sun Fire
667 2200/4150/4450 with ELOM.
668 nochecksumcheck - This workaround flag will tell FreeIPMI to not check
670 the checksums returned from IPMI command responses. It works around
671 systems that return invalid checksums due to implementation errors, but
672 the packet is otherwise valid. Users are cautioned on the use of this
673 option, as it removes validation of packet integrity in a number of
674 circumstances. However, it is unlikely to be an issue in most situa‐
675 tions. Those hitting this issue may see "connection timeout", "session
676 timeout", or "password verification timeout" errors. On IPMI 1.5 con‐
677 nections, the "noauthcodecheck" workaround may also needed too. Issue
678 observed on Supermicro X9SCM-iiF, Supermicro X9DRi-F, and Supermicro
679 X9DRFR.
680 idzero - This workaround flag will allow empty session IDs to be ac‐
682 cepted by the client. It works around IPMI sessions that report empty
683 session IDs to the client. Those hitting this issue may see "session
684 timeout" errors. Issue observed on Tyan S2882 with M3289 BMC.
685 unexpectedauth - This workaround flag will allow unexpected non-null
687 authcodes to be checked as though they were expected. It works around
688 an issue when packets contain non-null authentication data when they
689 should be null due to disabled per-message authentication. Those hit‐
690 ting this issue may see "session timeout" errors. Issue observed on
691 Dell PowerEdge 2850,SC1425. Confirmed fixed on newer firmware.
692 forcepermsg - This workaround flag will force per-message authentica‐
694 tion to be used no matter what is advertised by the remote system. It
695 works around an issue when per-message authentication is advertised as
696 disabled on the remote system, but it is actually required for the pro‐
697 tocol. Those hitting this issue may see "session timeout" errors. Is‐
698 sue observed on IBM eServer 325.
699 endianseq - This workaround flag will flip the endian of the session
701 sequence numbers to allow the session to continue properly. It works
702 around IPMI 1.5 session sequence numbers that are the wrong endian.
703 Those hitting this issue may see "session timeout" errors. Issue ob‐
704 served on some Sun ILOM 1.0/2.0 (depends on service processor endian).
705 noauthcodecheck - This workaround flag will tell FreeIPMI to not check
707 the authentication codes returned from IPMI 1.5 command responses. It
708 works around systems that return invalid authentication codes due to
709 hashing or implementation errors. Users are cautioned on the use of
710 this option, as it removes an authentication check verifying the valid‐
711 ity of a packet. However, in most organizations, this is unlikely to be
712 a security issue. Those hitting this issue may see "connection time‐
713 out", "session timeout", or "password verification timeout" errors.
714 Issue observed on Xyratex FB-H8-SRAY, Intel Windmill, Quanta Winter‐
715 fell, and Wiwynn Windmill.
716 intel20 - This workaround flag will work around several Intel IPMI 2.0
718 authentication issues. The issues covered include padding of usernames,
719 and password truncation if the authentication algorithm is HMAC-
720 MD5-128. Those hitting this issue may see "username invalid", "password
721 invalid", or "k_g invalid" errors. Issue observed on Intel SE7520AF2
722 with Intel Server Management Module (Professional Edition).
723 supermicro20 - This workaround flag will work around several Supermicro
725 IPMI 2.0 authentication issues on motherboards w/ Peppercon IPMI
726 firmware. The issues covered include handling invalid length authenti‐
727 cation codes. Those hitting this issue may see "password invalid" er‐
728 rors. Issue observed on Supermicro H8QME with SIMSO daughter card.
729 Confirmed fixed on newerver firmware.
730 sun20 - This workaround flag will work work around several Sun IPMI 2.0
732 authentication issues. The issues covered include invalid lengthed hash
733 keys, improperly hashed keys, and invalid cipher suite records. Those
734 hitting this issue may see "password invalid" or "bmc error" errors.
735 Issue observed on Sun Fire 4100/4200/4500 with ILOM. This workaround
736 automatically includes the "opensesspriv" workaround.
737 opensesspriv - This workaround flag will slightly alter FreeIPMI's IPMI
739 2.0 connection protocol to workaround an invalid hashing algorithm used
740 by the remote system. The privilege level sent during the Open Session
741 stage of an IPMI 2.0 connection is used for hashing keys instead of the
742 privilege level sent during the RAKP1 connection stage. Those hitting
743 this issue may see "password invalid", "k_g invalid", or "bad rmcpplus
744 status code" errors. Issue observed on Sun Fire 4100/4200/4500 with
745 ILOM, Inventec 5441/Dell Xanadu II, Supermicro X8DTH, Supermicro X8DTG,
746 Intel S5500WBV/Penguin Relion 700, Intel S2600JF/Appro 512X, Quanta
747 QSSC-S4R/Appro GB812X-CN, and Dell C5220. This workaround is automati‐
748 cally triggered with the "sun20" workaround.
749 integritycheckvalue - This workaround flag will work around an invalid
751 integrity check value during an IPMI 2.0 session establishment when us‐
752 ing Cipher Suite ID 0. The integrity check value should be 0 length,
753 however the remote motherboard responds with a non-empty field. Those
754 hitting this issue may see "k_g invalid" errors. Issue observed on Su‐
755 permicro X8DTG, Supermicro X8DTU, and Intel S5500WBV/Penguin Relion
756 700, and Intel S2600JF/Appro 512X.
757 ipmiping - This workaround option will inform ipmipower to use IPMI
759 based ping packets instead of RMCP ping packets. Some motherboards have
760 been observed to not implement RMCP ping/pong support despite being re‐
761 quired by the IPMI specification. Issue observed on Intel Windmill,
762 Quanta Winterfell, and Wiwynn Windmill.
763 No IPMI 1.5 Support - Some motherboards that support IPMI 2.0 have been
765 found to not support IPMI 1.5. Those hitting this issue may see "ipmi
766 2.0 unavailable" or "connection timeout" errors. This issue can be
767 worked around by using IPMI 2.0 instead of IPMI 1.5 by specifying
768 --driver-type=LAN_2_0. Issue observed on a number of HP and Supermicro
769 motherboards.
770
772 Upon successful execution, exit status is 0. On error, exit status is
773 1.
774 If multiple hosts are specified for communication, the exit status is 0
776 if and only if all targets successfully execute. Otherwise the exit
777 status is 1.
778 When operating in interactive mode, the exit value will be based on the
780 last power operation executed.
781
783 On older operating systems, if you input your username, password, and
784 other potentially security relevant information on the command line,
785 this information may be discovered by other users when using tools like
786 the ps(1) command or looking in the /proc file system. It is generally
787 more secure to input password information with options like the -P or
788 -K options. Configuring security relevant information in the FreeIPMI
789 configuration file would also be an appropriate way to hide this infor‐
790 mation.
791 In order to prevent brute force attacks, some BMCs will temporarily
793 "lock up" after a number of remote authentication errors. You may need
794 to wait awhile in order to this temporary "lock up" to pass before you
795 may authenticate again.
796 IPMI specifications do not require BMCs to perform a power control op‐
798 eration before returning a completion code to the caller. Therefore,
799 it is possible for ipmipower to return power status queries opposite of
800 what you are expecting. For example, if a "power off" operation is
801 performed, a BMC may return a successful completion code to ipmipower
802 before the "power off" operation is actually performed. Subsequent
803 power status queries may return "on" for several seconds, until the BMC
804 actually performs the "power off" operation.
805
807 Report bugs to <freeipmi-users@gnu.org> or <freeipmi-devel@gnu.org>.
808
810 Copyright (C) 2007-2015 Lawrence Livermore National Security, LLC.
811 Copyright (C) 2003-2007 The Regents of the University of California.
812 This program is free software; you can redistribute it and/or modify it
814 under the terms of the GNU General Public License as published by the
815 Free Software Foundation; either version 3 of the License, or (at your
816 option) any later version.
817
819 freeipmi.conf(5), freeipmi(7), ipmi-config(8), ipmi-oem(8)
820 http://www.gnu.org/software/freeipmi/
822ipmipower 1.6.10 2022-08-31 ipmipower(8)