CISC AS/400 LIC Tapes: Difference between revisions
(Expanded) |
(→Weblinks: Linkfix) |
||
| (5 intermediate revisions by the same user not shown) | |||
| Line 9: | Line 9: | ||
<blockquote>'''Note!''' It appears that the MULIC tape is needed only when the load source (aka: boot) disk is exchanged. A reinstall of LIC and OS does erase all of the disk's contents '''but''' MULIC. More evidence is required to support this claim, though. It has worked for a P02, and P03, so far. Sector-based backups of at least the load source disk is advised to not lose the MULIC part of a disk, along with the "personalized" parameters of the individual machine — aka. serial number.<p> | <blockquote>'''Note!''' It appears that the MULIC tape is needed only when the load source (aka: boot) disk is exchanged. A reinstall of LIC and OS does erase all of the disk's contents '''but''' MULIC. More evidence is required to support this claim, though. It has worked for a P02, and P03, so far. Sector-based backups of at least the load source disk is advised to not lose the MULIC part of a disk, along with the "personalized" parameters of the individual machine — aka. serial number.<p> | ||
It's very likely that changing *ULIC from external sources forces a change of the system's serial number, and thus an invalidation of the system password | It's very likely that changing *ULIC from external sources forces a change of the system's serial number, and thus an invalidation of the system password.</blockquote> | ||
== LIC == | == LIC == | ||
| Line 64: | Line 64: | ||
So, from the above list, it can be devised which models do not require MULIC/FULIC. For example, P03 models with processor 2115 and above need a tape, while P03 machines with CPU type 2114 do not. | So, from the above list, it can be devised which models do not require MULIC/FULIC. For example, P03 models with processor 2115 and above need a tape, while P03 machines with CPU type 2114 do not. | ||
== About the system Password == | |||
Steve Bisel makes it very clear in his book that the system password mechanism has been created from V2R2 onwards to make sure IBM gets paid its proposed license fees: Some software licenses are more expensive when run on more capable hardware. | |||
So far, this password is most likely calculated from system type, system serial number, and probably more information. This is being fed into kind of a hashing algorithm and outputs the magic string. Up to today, it can only be calculated by IBM.<ref>Latest experiences show that probably even IBM is no longer capable of generating passwords for just 30 year old machinery…</ref> | |||
* There is no OS provided way to view the current system password.<ref>The first release including the system password functionality — V2R2 — provided the possibility to print (not view!) the system password through the WRKHDWPRD command. This functionality was only briefly available for migration purposes, and was removed in V2R3.</ref> IBM stated that it is only needed once at install time — being provided on paper with the newly shipped system — and then never again, thus there is no need to know it. So, if it is invalidated, then it's for a reason and you need a new one generated, anyway. | |||
* Upon installation time of a machine, a system password is generated and "stored in the system". At least for the CISC models. | |||
* There is no concise documentation for all possible cases triggering a system password invalidation.<ref>On a model 400, adding a Token Ring IOA triggered a password invalidation for me. This is said to be a bug. No matter what it was, it was a major inconvenience to resolve.</ref> Cases might even be different between OS-Releases, and/or hardware involved. | |||
All in all, the less you tinker around, the less likely a sudden password invalidation will happen. What does '''not''' trigger an invalidation: | |||
* Adding or removing RAM (main storage). | |||
* Adding or removing DASD (disks). This holds true for MULIC-less RISC boxes' load source (boot disk). | |||
* Adding, moving, or removing feature (expansion) cards. | |||
While this is stemming from my personal experience with a few models, there's no guarantee that all models and releases will show a consistent behavior. It is said that especially V2 was hampered by bugs regarding System Password triggers. | |||
== See also == | == See also == | ||
| Line 71: | Line 87: | ||
== Weblinks == | == Weblinks == | ||
* [https://comp.sys.ibm.as400.misc.narkive.com/oTbiBhhr/looking-ibm-9401-p03 Looking IBM 9401-P03], Narkive Newsgroup Archive | * [https://comp.sys.ibm.as400.misc.narkive.com/oTbiBhhr/looking-ibm-9401-p03 Looking IBM 9401-P03], Narkive Newsgroup Archive | ||
* [https:// | * [https://www.ibm.com/resources/publications/OutputPubsDetails?PubID=SY44390200 AS/400 Service Functions V3], SY44-3902-00. See Appendix A for instructions about how to install LIC and how to deduce when MULIC is required. | ||
* Discussion about System Passwords on very computer:<ref>Note that some advice-givers confuse the system password with the QSECOFR user password, or license keys, for OS/400 releases V4 and newer.</ref> | |||
** [http://www.verycomputer.com/153_85a9beb296a124f7_1.htm Subject: System Password] | |||
** [http://www.verycomputer.com/153_58d5a1f3ef5378fd_1.htm Subject: System password] | |||
** [http://www.verycomputer.com/153_aa840c7ef6a45a77_1.htm Subject: How can I get the System Password Version?] | |||
** [http://www.verycomputer.com/153_aa8658dfd549509c_1.htm Subject: change system password] | |||
** [http://www.verycomputer.com/153_c178527a785f49aa_1.htm Subject: System password override?] | |||
== References == | |||
<references /> | |||
[[Category: System Administration]] | [[Category: System Administration]] | ||
Latest revision as of 13:08, 21 May 2023
Older, CISC AS/400 might have needed magnetic tapes with the hardware specific LIC saved to it.
These come in two flavors:
- Model Unique Licensed Internal Code, and
- Feature Unique Licensed Internal Code, associated with the newer Advanced series models, most often featuring a black case.
These LIC are very small binary blobs, less than 1 MB in size, and have hard-coded the system type (e. g. 9406), model (e. g. 200), and serial number.
Note! It appears that the MULIC tape is needed only when the load source (aka: boot) disk is exchanged. A reinstall of LIC and OS does erase all of the disk's contents but MULIC. More evidence is required to support this claim, though. It has worked for a P02, and P03, so far. Sector-based backups of at least the load source disk is advised to not lose the MULIC part of a disk, along with the "personalized" parameters of the individual machine — aka. serial number.
It's very likely that changing *ULIC from external sources forces a change of the system's serial number, and thus an invalidation of the system password.
LIC
According to Steve Bisel, the following RPQs were available for ordering from IBM:
| RPQ Number | AS/400 Model and Tape Media |
|---|---|
| S40302 | 9406 Bxx - Fxx models, MULIC tape, all media types except 8 mm on B models |
| S40347 | 9406 Bxx models, MULIC tape with 8 mm media |
| S40370 | 9402/9404 Bxx - Fxx models, MULIC QIC |
| S40371 | 9404 Bxx - Fxx models, MULIC ½" cartridge |
| S40372 | 9402/9404 Bxx - Fxx models, MULIC ½" reel |
| S40373 | 9402/9404 Bxx - Fxx models, MULIC 8 mm cartridge |
| 843819 | 9402/9404 3xx models, FULIC ½" cartridge |
| 843820 | 9402/9404 3xx models, FULIC ½" reel |
| 843818 | 9402/9404 3xx models, FULIC QIC |
| 843821 | 9402/9404 3xx models, FULIC 8 mm cartridge |
| 843817 | 9402/9404 3xx models, FULIC all media types |
The following text from around 2015 is literally taken from the weblink provided at the end of this article.
843916 - This RPQ is used to order a Bxx/Cxx All Models MULIC tape that has all the MULIC required for the B10, B20, B30, B35, B40, B45, B50, B60, B70, C06, C10, C20, and C25 models on one tape. MULIC is not required on the C04 model.
843917 - This RPQ is used to order a Dxx All Models MULIC tape that provides all the MULIC required for the D06, D10, D20, D25, D35, D45, D50, D60, D70, and D80 models on one tape. MULIC is not required on the D02 and D04 models.
843953 - This RPQ is used to order a Exx/Fxx All Models MULIC tape that provides all the MULIC required for the E04, E06, E10, E20, E25, E35, E45, E50, E60, E70, E80, E90, E95, F04, F06, F10, F20, F25, F35, F45, F50, F60, F70, F80, F90, F95, F97, P01, P02, 100, 135, and 140 models all on one tape. MULIC is not required on the A02, E02, and F02 models.
843954 - This RPQ is used to order an All Models FULIC tape that provides all the FULIC required for the P03, 10S, 20S, 200, 300, 310, 320, and 30S models. The All Models FULIC tape will have the FULIC for the 2010, 2031, 2032, 2040, 2041, 2042, 2043, 2044, 2050, 2051, 2052, 2115, 2116, 2117, 2411, and 2412 processor feature codes. FULIC is not required for processor feature codes 2030 and 2114.
So, from the above list, it can be devised which models do not require MULIC/FULIC. For example, P03 models with processor 2115 and above need a tape, while P03 machines with CPU type 2114 do not.
About the system Password
Steve Bisel makes it very clear in his book that the system password mechanism has been created from V2R2 onwards to make sure IBM gets paid its proposed license fees: Some software licenses are more expensive when run on more capable hardware.
So far, this password is most likely calculated from system type, system serial number, and probably more information. This is being fed into kind of a hashing algorithm and outputs the magic string. Up to today, it can only be calculated by IBM.[1]
- There is no OS provided way to view the current system password.[2] IBM stated that it is only needed once at install time — being provided on paper with the newly shipped system — and then never again, thus there is no need to know it. So, if it is invalidated, then it's for a reason and you need a new one generated, anyway.
- Upon installation time of a machine, a system password is generated and "stored in the system". At least for the CISC models.
- There is no concise documentation for all possible cases triggering a system password invalidation.[3] Cases might even be different between OS-Releases, and/or hardware involved.
All in all, the less you tinker around, the less likely a sudden password invalidation will happen. What does not trigger an invalidation:
- Adding or removing RAM (main storage).
- Adding or removing DASD (disks). This holds true for MULIC-less RISC boxes' load source (boot disk).
- Adding, moving, or removing feature (expansion) cards.
While this is stemming from my personal experience with a few models, there's no guarantee that all models and releases will show a consistent behavior. It is said that especially V2 was hampered by bugs regarding System Password triggers.
See also
- Copying disks with Linux
- Steve Bisel, AS/400 Power Tips and Techniques, First ed. June 1995 covering OS/400 V3R1, ISBN 1-883884-13-6
Weblinks
- Looking IBM 9401-P03, Narkive Newsgroup Archive
- AS/400 Service Functions V3, SY44-3902-00. See Appendix A for instructions about how to install LIC and how to deduce when MULIC is required.
- Discussion about System Passwords on very computer:[4]
References
- ↑ Latest experiences show that probably even IBM is no longer capable of generating passwords for just 30 year old machinery…
- ↑ The first release including the system password functionality — V2R2 — provided the possibility to print (not view!) the system password through the WRKHDWPRD command. This functionality was only briefly available for migration purposes, and was removed in V2R3.
- ↑ On a model 400, adding a Token Ring IOA triggered a password invalidation for me. This is said to be a bug. No matter what it was, it was a major inconvenience to resolve.
- ↑ Note that some advice-givers confuse the system password with the QSECOFR user password, or license keys, for OS/400 releases V4 and newer.