Date: Sun, 01 Mar 1998 16:30:28 -0700 (MST) From: Joe Doupnik Subject: FAQ is searchable To: FLOYD@DIRECT.CA Floyd, I spent this weekend finding a workable web search engine to put upon netlab1 so the FAQ could be culled. Well, it's done, not perfect but usable. If you web browse to netlab1.usu.edu then at the bottom of the home page are four styles for searching. The database is only the FAQ and my short Win95 supplement. While searching for searchers I discovered many are junk or abandoned etc. Finally one appeared that actually worked, and I then proceeded to figure out how it worked and beat it into shape. It's a freebee edition, but limited to 100K source files. Joe D. --------- Date: Tue, 03 Mar 1998 10:42:32 -0700 (MST) From: Joe Doupnik Subject: Re: FAQ is searchable To: floyd@direct.ca >I surfed your FAQ-search link...quite cool, "works as expected". > >Questions: (1) does it run on NetWare?, (2) what is the link to >it, so others can try it out on their FAQ mirror page?, (3) can I >help in the implementation by doing something once so that others >don't have to repeatedly do something?, (4) is that <100,000 byte >files, or <100,000 files? ------- Netlab1 is a UnixWare machine, so it's Unix. For 1.5 years Novell has promised to outfit me with a working NW based web server. Alas, NFS for NW is needed and has not been satisfactory. You may recall I'm the guy who caused that product to be rewritten, and this month Novell is sending up one of the design engineers for a day's review place. So for now it's Unix rather than NW. The search engine runs on the same machine. If one chooses say Webinator style and Help then the vendor reference (www.thunderstone.com) is revealed. This means other sites can grab the same material, assuming their machinery is supported, and then grab my rendition of the search pages, and there we are. The netlab1 "home page" item is readily seen in source form. The limit on free usage is 100,000 files (what it terms pages). No limit on size of individual files. The FAQ consumes about 155 files. I don't think there is much more we need do. There are a few stray IP addresses that don't parse, but that's expected. Joe D. ------------------------------ Date: Tue, 10 Mar 1998 21:02:44 -0700 From: Joe Doupnik Subject: FAQ searching at netlab1, a clarification Some queried me this evening about that mentioned FAQ search tool, saying they didn't see it. This is on netlab1.usu.edu (ok, in web-speak, http://netlab1.usu.edu/). I suspect the person did what many do: created a history entry which included a trailing novell.faq/ to the site name. Well, the FAQ file set itself doesn't have a search button per se. The reason is fairly obvious: there are many sites round the world sharing the same files from Floyd. Thus search engines would be different (if they existed at all) and the pointer expression etc would differ too. So no such button in the FAQ docs themselves (which live in my directory /novell.faq/). Instead things are actually simpler (have to be for me to cope with this stuff). On the "home page" of netlab1 (http://netlab1.usu.edu/ and nothing else) is that Search selection menu, near the bottom of the page. I like Webinator style as fanciest, many perfer AltaVista style. Whatever. Golly, this web stuff is getting harder to use by the week, what with growth and embellishments and screens too full to show everything. In any case, I do welcome suggestions on improvments that are within my capabilities here. Oh yes, all this works ok with text mode Lynx too. Joe D. P.S. My grad networking class made two suggestions: please index all the RFCs, and the Unix man pages too, thanks. Alas, those indices are large, so the answer to them was sorry, no. ------------------------------ From: Subject: german Novell FAQ Date: Thu, 25 Dec 1997 13:46:28 +0100 I don't know if you want to include a link to a german Novell FAQ (not the one from de.comp.sys.novell), but here are the URLS: The Novell FAQ itself: http://pandora.inline.de/novell/faq/ Some information about the FAQ and download addresses (in german): http://pandora.inline.de/novell/faq.htm Just a list of links to English FAQs: http://pandora.inline.de/novell/faq_e.htm This Novell FAQ is updated regularly every month and is originally based on mails from members of the Fidonet Echo (Newsgroup) NOVELL.GER. H7. Register memory ("Cache memory alloc. out of avail. memory" msg) SOLUTION EISA machines should be configured to use AUTO REGISTER MEMORY, but in the case of ISA machines or EISA machines that are not configured correctly, the following kludge will solve the problem. Every machine that requires the REGISTER MEMORY command, uses more than 16MB of RAM, and has large disk storage, will exhibit problems such as this. This is not an issue in machines that do not require the REGISTER MEMORY command (such as the COMPAQ SystemPro, PS/2 MODELs 90, 95, HP Vectra etc.) because at boot up you see all of the RAM. ... or if you applied the LOADER patch from 312PTx.EXE. This allows even non-EISA machines to access all memory without playing with different AUTOEXEC files and REGISTER MEMORY. Most new machines with PCI bus work without any problem. Nevertheless, some old machines still need this patch, and all machines running with NW 3.11, since AFAIK the LOADER patch doesn´t resolve the memory problem there. There, .... ------ file server name HOPE ipx internal net 9088 register memory 1000000 1000000 load isadisk port=1F0 int=E mount SYS autoboot ------ instead of "autoboot" use the line "sys:system/autoexec.ncf" to load the original AUTOEXEC.NCF (without FILE SERVER NAME and INTERNAL IPX), which has the advantage that you can still use SYSCON and INSTALL.NLM to edit the AUTOEXEC.NCF. In theory, manually purging files on NetWare should not be necessary. Experience has shown, however, that there are advantages to purging. [...] To automatically purge files from a file server, but not to purge all in a given directory with FLAGDIR ... P, there a some PURGE.NLMs available for example on http://pandora.inline.de/novell/files/util/nlm/ (with links to 2 mirror servers in germany and austria) The descriptions on the webpage are in german, but most of the files (and accompaning docs) are in english. You can use this URL in the Novell FAQ, too, if you think that the utilities listed are worth mentioned. (perhaps better http://pandora.inline.de/novell/files/) I would prefer to use the PURGE command from TBOXx.EXE from Novell. Since this Tool is quite new, please use always the newest version, most of all since the former versions of TBOX caused the file server to abend when you used PURGE. ------------------------------ From: Subject: A FAQ suggestion Date: Sat Feb 17 20:33:02 1996 >$I am setting up a small network using Novell's two-user CD and am >$trying to get by without having to buy a hub at this point. Does >$anyone know on what numbers the wires cross on an RJ45 to make this work? > > Your crossover cable will be wired: > > 1 ---- 3 > 2 ---- 6 > 3 ---- 1 > 6 ---- 2 $ I read your recent posts about netware 3.X not $allowing the use of the the 0 and 1 subnets. I've always believed $that myself, but when I spoke to a TCP/IP person at Novell the other $day, she told me that the latest 3.12 TCP/IP update would let you use $the 0 subnet if you wanted to. I haven't tried it, but I just wanted $to relay it to you. I know that the latest TCPIP.NLM for 4.10 allows this; actually, if memory serves, the TCPIP.NLM which ships with 4.10 may allow it, too. The latest TCPIP.NLM for 3.1x has some features of the 4.10 version but not all of them, and I don't know if this is one of them. >>>I've rewritten the following section, based on the earlier version. E.5.1 Why It Doesn't Always Work As Expected Internet standards are defined by papers known as RFCs (Request For Comments). Novell's TCP/IP implementation in NetWare versions prior to 4.10 is based on RFC950, which states that no portion of a node's address may consist of all zeroes or of all ones. It has become common practice in recent years to use all-zeroes and all-ones subnets; the latest TCP/IP update for NetWare 3.12, and the TCP/IP stack included in NetWare 4.10, allows the all-zeroes subnet but continues to disallow all ones. The remainder of this answer assumes that neither the all-zeroes nor the all-ones subnets may be used; this is the safest assumption to make in any case because there may be other devices on your network which also follow RFC950. You can logically break down an address into two components where subnetting is not used (the network portion, defined by the address' class; and the node portion), and three components where subnetting is used (the network address; the subnetwork address, which is the difference between the subnetwork mask and the normal mask used for that address class; and the node address). By RFC950, none of those three portions may consist of all zero bits or all one bits. Let's look at an example. Let's say we're subnetting a class B address, 166.166.0.0, into a number of class C-sized subnetworks. We would do this with a subnetwork mask of 255.255.255.0. This yields 254 networks, 166.166.1.0 through 166.166.254.0. We cannot use 166.166.0.0, as the subnetwork portion (the third octet) is all zeroes; similarly, we cannot use 166.166.255.0, as the subnetwork portion is all ones. A more common example, though, is the subnetting of a class C address. Let's use 200.200.200.0 as an example, and say that we want two subnets. Using 255.255.255.128 would result in two illegal subnets - 200.200.200.0, in which the subnetwork portion (the first bit of the last byte 0) is all zeroes, and 200.200.200.128, in which the subnetwork portion is all ones. Therefore, we can't use it. To get two subnets, we must use the mask 255.255.255.192, which gives two valid subnets - 200.200.200.64 (with 62 nodes numbered .65 through .126 inclusive) and 200.200.200.128 (with 62 nodes numbered .129 through .190 inclusive). For subnets of a class C address, your choices are: Mask Number of Subnets Size of each Subnet 255.255.255.0 1 (not subnetted) 254 nodes 255.255.255.128 0 (not allowed) N/A 255.255.255.192 2 62 nodes 255.255.255.224 6 30 nodes 255.255.255.240 14 14 nodes 255.255.255.248 30 6 nodes 255.255.255.252 62 2 nodes 255.255.255.254 0 (not allowed) N/A In Netware 4.10, 255.255.255.128 is allowed, but yields one valid subnet of 126 nodes (so it's pointless), and all lines below it yield one more valid subnet than is indicated above. And a brand new answer on subnets, particularly those of variable sizes: First, a definition. RIP is the traditional routing information protocol used by IP routers to exchange information on networks. RIP predates the concept of subnets, and is incapable of expressing subnetting information. Two newer protocols, RIP II and OSPF, are much more advanced. More on these protocols as we go along. Normally, all subnets of a given address are of the same size. RIP, in a subnetted environment, is therefore usually written to assume that whatever the subnet mask is on a particular interface, it's the same for all other subnets of that interface. This is a bit of a kludge, but it works - if, and only if, this assumption holds true. You can find people who will argue that RIP should never be used, or that it should never be used in a subnetted situation, or that if you're careful, you're OK with it in this case ... There may be cases in which you wish to break a single network address into subnets of different sizes. This is known as using a variable length subnet mask (VLSM). Prior to NetWare 4.10, Novell's TCP/IP stack did not permit this at all. In NetWare 4.10, it is permitted. However, RIP is incapable of handling VLSM. If you need to use a routing information protocol in a VLSM environment, use RIP II or OSPF; these can be configured using INETCFG.NLM. You may not actually need a routing protocol in many cases. Routing protocols are used by router to exchange information about how each other is configured. If you manually configure each router which needs to know, then routing protocols are not needed. This is known as static routing. For example, consider the following network: Internet ----- Router ----------------- Server ---------------- Network A Network B The rest of the world needs to know that networks A and B are both reached through the router. The router needs to know that network B is reached through the server. And the server needs to know that the rest of the world is reached through the router. This information does not change dynamically, and is relatively simple; it can easily be handled by static routes. In fact, if something on the network doesn't handle RIP II or OSPF, and for some reason you've had to use VLSM, then static routing is probably your only option. In this simple example, the need for VLSM is unlikely - but it's not difficult to require VLSM by adding a couple of additional networks, of varying sizes, and having to fit it all into one class C address. >>>Here's an addition to this answer: G.3 Getting garbage in printouts There's also a bug in some versions of RPRINTER which causes a single garbage character to be sent to the printer when RPRINTER is first run. Try upgrading to a current RPRINTER.EXE to cure this. >>>We should mention SYNC.NLM in the following answer. I don't >>>recall exactly where it comes from, though ... H.24 How can I keep my server on time? [add after the usual text for this one:] Another option, if you have one NetWare server which keeps good time or which uses an external time source, is to keep your other servers in sync with it using SYNC.NLM. This NLM periodically polls a given server (the polling interval is adjustable) to get the correct time, and corrects the local server's time if it's not right. It can be configured to take timezone differences into account. >>>A minor adjustment to: H.27 Advantages of purging files In theory, manually purging files on NetWare should not be necessary. Experience has shown, however, that there are advantages to purging. Purging files that are temporary in nature allows other more essential deleted files to have a longer "deleted but not purged" (ie. still salvageable) life. Also, any \TEMP directories, and all print queue directories (those funny numbered ones under SYS:system), should be marked with FLAGDIR as "purgeable" e.g. FLAGDIR Z:\SYSTEM\10000001 PURGE [add the following to the above paragraph] In NetWare 4.x, FLAGDIR was rolled into FLAG, so you'd use FLAG Z:\SYSTEM\10000001 P >>>A correction: I.4.4 Packet tuning [...] Another tip regards NCP searches. I found that the Requester performs better on searching through directories when you increase the "Maximum Outstanding NCP Searches" console set parameter from the standard value of 51 to say, 200 or 300. On NetWare 3.x servers, use SERVMAN NLM to play with it. On NetWare 4.x servers you can use SERVMAN NLM to play with these parameters. [replace the "On NetWare 3.x servers" sentence with] On NetWare 3.x servers, you'll have to adjust this manually with SET commands from the console, and incorporate the end result into your AUTOEXEC.NCF. >>>And we should update the CD-ROM material ... please replace the N.1 question, and the first paragraph of the answer, with: N.1 How do I get to see a CD drive on my Netware v3.12 or higher server? Only SCSI CD-ROMs can be mounted by NetWare straight out of the box. There is an update, CDUP2.EXE, which adds support for ATAPI drives. Treat this update with caution, as it may not be entirely stable. Also, ATAPI CD-ROMs have no place in a serious fileserver, just as IDE hard drives don't. And be forewarned that many older CD-ROM drives which appear to be IDE-like may _not_ be ATAPI drives, but may instead use some proprietary IDE-like interface which is not supported by CDUP2. >>>Replace the first paragraph of answer N.2 with: When mounting a CD-ROM on a Netware v3.12 or 4.0x server, CPU utilization may go to 100% for a long time. This is normal for the original 3.12 CDROM.NLM, but has been at least partially cured in later releases. You should upgrade to the current release (NetWare 4.10 users, as of this writing, have the current release; all others need CDROM3.EXE). ------------------------------ Date: Wed, 14 Feb 1996 11:29:04 -0800 From: steve@sdesign.com (Stephen Herzog) Subject: Request for updates to the NetWare FAQ. 3) It may be nice to combine section G.1 and section G.12 as they cover the same topic. G.1 Printing to/from Unix systems G.12 Transfering print jobs between NetWare and Unix print queues ------------------------------ Date: Mon, 4 Dec 1995 09:29:49 GMT From: Peter McGarvey Subject: Re: conlog.nlm >I use conlog.nlm to capture information from the console screen. I >love it except for one thing: conlog writes to the same file (console.log) >each time. So, if my server crashes I lose whatever information is >contained in the file because of the overwrite. Which sort of defeats >the purpose of having it. Does anyone know of parameters I can set >to prevent this from happening? CONLOG maximum=1024 save=SYS:\ETC\OLD.LOG Now all you have to do is to rename OLD.LOG each time it gets created. I personally send it to a separate directory with the name YYMMDD.LOG --------- From: "Paul Massue-Monat" Date: Tue, 16 Jan 1996 12:51:23 EST Subject: Re: ConLog -- Options or alternatives? I agree with you on the difficulty of conlog. Maybe eventmon.nlm could be of help. It records events but I'm not sure it would record "problems". I haven't tried it myself since I just learned of the nlm one week ago and I didn't have a chance to download it. I discovered it by arriving at this edinburgh web site: http://mft.ucs.ed.ac.uk/novell/techsup/freenlms.htm (they list about a dozen free useful NLMs). From here, it's allways difficult to get to Great Britain: the line is allways slow. Eventually, I'll use archie to find the file eventm.zip closer to me (netlab ? or the netlab mirror in Ottawa at nrc ?). The file eventm,.zip is supposed to contain eventmon.nlm P.S.: Joe is right to compain about the web but I sure would like a search tool at netlab. Those "whatIS" files are so difficult to use, I would love it if he would put up a search facility a la web (just like there is at novell.com). In fact, novell.com is very easy to use these days with a web browser; example: all updates for nw4.1 are in one directory called nw/os/410 or somesuch; the same for 3.11, etc. --------- Date: Fri, 19 Jan 1996 07:21:18 -0800 From: mpierce@danasoft.com (Michael Pierce) Subject: EVENTMON We no longer own this product as we sold it to Citadel System (number 800-962-0701) about a year and a half ago. If you have a specific question I can try to answer it, but we no longer own the rights to the product. DanaSoft, Inc. (703) 257-4650 http://www.danasoft.com --------- Date: Thu, 15 Feb 1996 13:44:30 +0100 From: "David W. Hanson" Subject: Re: CONLOG question [Someone] wanted to know conlog command line switches. ...from the Dyna-Text: LOAD CONLOG [FILE=log filename] [SAVE=backup filename] [MAXIMUM=max file size] [ENTIRE=YES] | HELP Parameter - Use to FILE=log filename - Specify the log file you want to use if you don't want CONLOG to write to the default log file (CONSOLE.LOG). The log filename parameter is used with SFT III servers, whose three engines must each write to a different log file. SAVE=backup file name - Specify the filename in which you want to save the preexisting console log. This parameter enables AUTOEXEC.NCF to load CONLOG without overwriting the previous console log. Use this parameter if you want to examine the old console log for troubleshooting purposes. MAXIMUM=max file size - Specify the maximum size, in kilobytes, of the log file. If the log file exceeds the max file size, it is deleted---or renamed if backup filename was specified. In either case, logging starts over with an empty log file. ENTIRE=YES - Capture the lines already on the screen at the time CONLOG is loaded. Without this parameter, CONLOG starts logging with the next output to the console after it is loaded. HELP - View a description of each CONLOG command parameter ------------------------------ Date: Thu, 08 Feb 1996 10:43:23 -0800 From: Marcus Williamson <71333.1665@Compuserve.com> Subject: NETW4X.FAQ You might like to mention two programs which may help in sections H.25.2 and H.7 of the FAQ. Memory calculation : SMEM.EXE REGISTER MEMORY calculation : REGMEM.EXE Both these, and NETW4X.FAQ, previously only in the CompuServe forums, are now available at : http://ourworld.compuserve.com/homepages/marcus_williamson I'd appreciate if you could add this to your FAQ references. The file includes many FAQs have been gathered from the CompuServe forums, Usenet and customer contact. Many of the Qs & As do not appear anywhere else in existing FAQ material. ------------------------------ Date: Fri, 19 Jan 1996 15:05:23 -0500 From: bhatia@chelsea.ios.com (Deepak Bhatia, Master CNE) Subject: Possible Addition to the FAQ How about a new section for the FAQ - Online Learning Centers: Cyber State University: ----------------------- Offer online virtual classes which combine Video, interactive chats with students, reading from the CNE guide for NetWare 4.1, testing and hands on lab exercises. To check this out call (510) 253-TREK MOLI: ----- The Microsoft Online Institute is an online interactive learning and information resource available on MSN, the Microsoft Network online service, to Windows 95 users. Using a campus paradigm, MOLI provides online access to learning materials, instructor expertise, product information, developer articles, user forums, and other resources for Microsoft product and technology information. In addition, MOLI provides information about all Microsoft training options. For more details go to Microsoft page on the World Wide Web at: http://www.microsoft.com/moli ------------------------------ Date: Wed, 27 Dec 1995 00:24:59 +1000 From: Richard Phillips Subject: H.37 Negative numbers from CHKVOL -- yup, time to worry H.37 Negative numbers from CHKVOL -- yup, time to worry -> With Netware 4.x "chkvol" is replaced with "ndir /vol", and this has been known (by me at least) to show some weird figures with 4.02 and 4.1 so you may not actually have to worry or run vrepair. If this worries you, get the latest version of ndir.exe from netwire at "ftp.novell.com/pub/updates/nwos/nw410/41ndir.exe" which appears to have fixed these problems. Although it's in the 4.1 update directory, this works with 4.02 as well as 4.1, and should work with all other older versions of 4.x although I'vce only run it with 4.02 and 4.1 Nwadmin has also been known to show weird figures, particularly with reference to compressed space and percentages - and "they're working on it". ------------------------------ Date: Tue, 26 Dec 1995 23:58:18 +1000 From: Richard Phillips Organization: McIntosh & Company Subject: H.6 Changing the number of licenced users on a Netware server H.6 Changing the number of licenced users on a Netware server Netware 2.x - requires a full "netgen" Netware 3.x - as already in FAQ, just replace server.exe Netware 4.x - the license is separate to the server.exe file. To increase or decrease the number of licenses is a matter of adding or deleting the license.mls file via install.nlm from the server console. With Netware 4.0x you may replace the current license via install.nlm, and with Netware 4.1 you may delete the previous license and/or add additional licenses. If you have two 50 user 4.1 licenses, for example, you can now create a 100 user server. If you have one 100 user license, however, you cannot split it into smaller units (at the moment) unless you can find someone to swap the license with for an equivalent number of smaller blocks of licenses. ------------------------------ From: Steven M. Dunn Subject: NetWare FAQ comments Date: Fri, 27 Oct 95 9:49:54 EDT >>G.8 Alternatives to PSERVER for more than 16 printers, etc. > > Again, NetWare 4 relieves this, by allowing PSERVER to handle > lots (255, if memory serves) of printers. OK, I was wrong. It's 256 :-) >>Alternatively, *FOR NETWARE 2.X SERVERS ONLY*, you can try the following: [...] > This isn't right; there are only two bindery files in 2.x. > Darn if I can remember 'em, though ... again, I *think* the Hacking > FAQ has 'em. NET$BIND.SYS and NET$BVAL.SYS are the two files for 2.x. >>H.24 How can I keep my server on time? [...] > 2. If you have one NetWare server which you keep set with the > correct time, you can have other 3.x servers follow its time using > SYNC.NLM (available from somewhere on Compuserve ... I'll try to > dig up the reference to it). ftp://ftp.novell.com/pub/netware/novuser/07/sync.zip; I expect that's the same directory it's in on CompuServe, too. RDATE.ZIP is also in this directory, for those who wish to sync their time to a known time source via TCP/IP. ------------------------------ I also have a few proposals, it could be worth to mention in section E.7 (BOOTP and RARP): In LAN WorkPlace or -Group (and the TCPIP in VLM Client Kit) if you use 'ip_adress 0.0.0.0' in NET.CFG it will first look for a BOOTP-server, if no bootp server respond it will look for a RARP-server. In F.3 (Mercury): You could mention that Mercury also have a function as 'Mail Server' with SEND 'filename', INDEX, LOOKUP, .... it also can act as a (simple) 'List Server' with SUBSCRIBE, UNSUBSCRIBE, then you have to dedicate a PC running Pmail 24h, and use filtering to handle in/out going mail to the list. Steen Rahbek ------------------------------ : Does any one know where I can find a list of Netware API calls : declaration for Visual Basic if there is one? TIA! Joseph Devito has written some vb definitions to Netware. You can find them at ftp.shu.edu in /pub/nw41/nw41bas.exe. Bengt Bolinder ------------------------------