[OpenAFS] AFS reinstall
Hello, I am trying to get our primary fileserver (running all the file,db,volume,authentication processes) back up after the machine was hacked. The server was running RH8 with OpenAFS 1.2.10. The system disks were wiped after backing up /usr/afs, and CentOS4(RHEL4) was installed. Using all the old config files and vice partitions, is there a problem with installing the latest stable version(1.4.1) OpenAFS rpms. Thanks in Advance, Gabe ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
Re: [OpenAFS] /usr/afs/etc/KeyFile from krb4?
There is currently only 1 server. Will generating a new KeyFile corrupt data or the database, or somehow lose user accounts. Can someone give me a quick rundown on how to do this, it's been quite a while. Also, if there is a way to convert the old KeyFile and db to something usable, any pointers would be much appreciated. Thanks, Gabe "Christopher D. Clausen" <[EMAIL PROTECTED]> wrote: Gabe ListAccount wrote:> Hello,>I have a server that was hacked, and thus a new OS (CentOS4) was> installed. I setup OpenAFS 1.4 , openafs-krb5-1.4.1 was installed. I> dropped the old db files as well as the KeyFile into their respective> directories. I don't think this was appropriate. How do I convert the> old KeyFile and db (from OpenAFS 1.2.10) to be compatble with krb5?Uhh, well, if your server was hacked you likely do not want to the use the old KeyFile and instead generate a new one. You would need to add the updated key to all AFS servers in your cell and you should remove the old key as quickly as possible.In thet past people have used something called the Kerberos 5 Migration Kit to go from AFS kaserver to Kerberos 5. I'm not sure if that is still the recomended thing to do or not though. I thought that at least MIT Kerberos 5 could read the older Kerberos db file from kaserver.<-- Christopher D. Clausen[EMAIL PROTECTED] SysAdminLove cheap thrills? Enjoy PC-to-Phone calls to 30+ countries for just 2�/min with Yahoo! Messenger with Voice. New Yahoo! Messenger with Voice. Call regular phones from your PC and save big.
[OpenAFS] /usr/afs/etc/KeyFile from krb4?
Hello, I have a server that was hacked, and thus a new OS (CentOS4) was installed. I setup OpenAFS 1.4 , openafs-krb5-1.4.1 was installed. I dropped the old db files as well as the KeyFile into their respective directories. I don't think this was appropriate. How do I convert the old KeyFile and db (from OpenAFS 1.2.10) to be compatble with krb5? Thanks in Advance, Gabe Love cheap thrills? Enjoy PC-to-Phone calls to 30+ countries for just 2¢/min with Yahoo! Messenger with Voice.
[OpenAFS] MS Access and OpenAFS
Has anyone used an MS Access database from within OpenAFS? Does it work with multiple users? I'd heard some talk about AFS not supporting byte-range locking, and wasn't sure how MS Access fit into this. Thanks, Gabe ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] UBIK error installing additional database server on 1.2.10
Hello, I'm getting an error after adding database functionality to a 2nd fileserver with OpenAFS 1.2.10. The error"Waiting for quorum election." fills up the logs on the server I was trying to add database server functionality to. I've read that this is a bug in 1.2.10. My question is if I add /usr/afs/db to the upclientetc process on the new db server, will this allow me to remove the db functionality from the first and ONLY db server. I looked at the files in the /usr/afs/db directory, and they are not the same. Also, is it safe to just scp that directory to the new server and run it as the dbserver with the original shut down. Thanks, --Gabe ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
Re: [OpenAFS] repeated failed volume move operation
Hi, Here's some more info from the logs. The command was run as [EMAIL PROTECTED] vos move sata.vol fs1 vicepc fs2 vicepc -local -verbose from the VolserLog on fs2(destination): Tue Mar 22 06:21:08 2005 trans 36 on volume 536870924 is older than 40560 seconds Tue Mar 22 06:21:20 2005 1 Volser: ReadVnodes: Restore aborted ...Many more Trans 37 on volume 536870924 is older than messages Tue Mar 22 06:49:09 2005 trans 37 on volume 536870924 is older than 1650 seconds Tue Mar 22 06:49:27 2005 1 Volser: Delete: volume 536870924 deleted from the VolserLog on fs1(source): Mon Mar 21 17:59:44 2005 1 Volser: Clone: Cloning volume 536870924 to new volume 536870954 ... Tue Mar 22 06:20:12 2005 1 Volser: DumpVolume: Rx call failed during dump, error 1492325122 Tue Mar 22 07:45:58 2005 1 Volser: Delete: volume 536870954 deleted Jeffrey Hutzelman wrote: Tue Mar 22 07:45:36 2005 trans 38 on volume 536870954 is older than 3420 seconds That's normal; it's just being chatty. Tue Mar 22 07:45:58 2005 1 Volser: Delete: volume 536870954 deleted Also normal; this is the move clone being deleted. VOLSER: Problems encountered in doing the dump ! This is src/volser/vsprocs.c:PrintError()'s rendition of VOLSERDUMPERROR. What it generally means is that the source volserver ran into an error writing into the network channel between the two servers. That is most likely the result of the call having been aborted on the destination server for some reason. Have you looked in the VolserLog on that server? -- Jeffrey T. Hutzelman (N3NHS) <[EMAIL PROTECTED]> Sr. Research Systems Programmer School of Computer Science - Research Computing Facility Carnegie Mellon University - Pittsburgh, PA ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] repeated ailed volume move operation
I looked through some of the old emails, and have seen this before. I couldn't find a solution The operation is started from fs2. I watch it, and it transfers a large amount of data. This volume is ~400GB . While checking up on it, it transferred at least 200GB. It has done this 3 times. The only thing I get in the Volserlog is : Tue Mar 22 07:45:36 2005 trans 38 on volume 536870954 is older than 3420 seconds Tue Mar 22 07:45:58 2005 1 Volser: Delete: volume 536870954 deleted There are many older than X seconds messages. I thought that was just because it was a large volume, and would thus take a long time. Is there some sort of timeout I need to bump up. vos move sata.vol fs1 vicepc fs2 vicepc -local -verbose Starting transaction on source volume 536870924 ... done Cloning source volume 536870924 ... done Ending the transaction on the source volume 536870924 ... done Starting transaction on the cloned volume 536870954 ... done Creating the destination volume 536870924 ... done Dumping from clone 536870954 on source to volume 536870924 on destination ...Failed to move data for the volume 536870924 VOLSER: Problems encountered in doing the dump ! vos move: operation interrupted, cleanup in progress... clear transaction contexts access VLDB move incomplete - attempt cleanup of target partition - no guarantee cleanup complete - user verify desired result Thanks, Gabe __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] fs process doesn't exit until I send a signal 9
I had to shut down one of my AFS servers to replace a disk. When I issue a "bos shutdown" command, all the processes seem to shutdown, except for the fs process. When I "vos status" the server, it says that the fileserver has been disabled, and the sub-status is "in the process of shutting down". Is there something else I need to do. I only seem to be able to shut it down by using "killall -9 filserver" or "kill -9" the fileserver process number. This of course causes the volumes to be salvaged. Is there something else that I need to do? How long should it take to shutdown the fs process. I waited about 5 minutes before using "kill -9". Below is /usr/afs/logs/BosLog after the vos shutdown command: Thu Feb 24 11:14:44 2005: fs:vol exited on signal 15 Thu Feb 24 11:14:44 2005: upserver exited on signal 15 Thu Feb 24 11:14:44 2005: buserver exited on signal 15 Thu Feb 24 11:14:44 2005: vlserver exited on signal 15 Thu Feb 24 11:14:44 2005: ptserver exited on signal 15 Thu Feb 24 11:14:44 2005: kaserver exited on signal 15 Thu Feb 24 11:19:12 2005: fs:file exited on signal 9 And also, the BosLog after a "bos restart fs" Thu Feb 24 11:32:22 2005: Server directory access is okay Thu Feb 24 11:37:16 2005: fs:salv exited with code 0 Thu Feb 24 13:04:55 2005: fs:vol exited on signal 15 Thu Feb 24 13:11:47 2005: fs:file exited on signal 9 Thu Feb 24 13:17:28 2005: fs:salv exited with code 0 --Gabe Castillo [EMAIL PROTECTED] ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] restoring volumes > 2GB
Correct me if I'm wrong, but the 2GB size limit applies to files correct? not volumes? I have volumes that are several hundred GB and have successfully restored them from tape. ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] Trying to recreate root.cell
Hello, I am trying to restore data from a crashed RAID. I have the data on a different fileserver running reiserfs. Is this the source of my problems? If I recreate a RAID array of the original size will this work? Is there a way to get this to work on the newer fileserver? Or am I totally out of luck because the location of the files based on the original inodes cannot be recreated. Help is greatly appreciated. This is ~.5TB of unbacked up data. If I can't exactly restore the volume in full is there a way to extract the files out of the /vicepX partition. Below is the output of the SalvageLog 01/06/2005 01:18:19 STARTING AFS SALVAGER 2.4 (./salvager) 01/06/2005 01:18:19 Starting salvage of file system partition /vicepa 01/06/2005 01:18:19 SALVAGING FILE SYSTEM PARTITION /vicepa (device=vicepa) 01/06/2005 01:18:21 ***Forced salvage of all volumes on this partition*** 01/06/2005 01:18:21 Recreating link table for volume 36870918. Unable to allocate link table inode for volume 536870918 (error = 17) 01/06/2005 01:18:21 SALVAGING OF PARTITION /vicepa COMPLETED Much Thanks, Gabe [EMAIL PROTECTED] __ Do you Yahoo!? Yahoo! Mail - Helps protect you from nasty viruses. http://promotions.yahoo.com/new_mail ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] re-installing a "lost" volume
Hello, I have had a data-loss issue on the RAID with the root AFS volume. I have recovered the data, and transferred it to a different fileserver. I am curious if I can modify the VLDB to tell AFS that the volume is on a different machine. Does this pose any potential problems (particularly because the volume isn't available when I have to modify the VLDB)? If this is unsuccessful is there a problem with modifying the VLDB back to its original setup? Thanks in advance, Gabe Castillo ___ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] Volume Management
Hello, Does anybody have any strategies for volume management. I run an AFS cell of a large size(~4TB) for a small number of users (~10 -20 users). People in my lab want an endless amount of space since most of the directories are used by groups of people . I was just curious to hear how other organizations manage their disk space. e.g. is it all in user-based volumes? , etc. Thanks, Gabe Castillo ___ OpenAFS-info mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] root.afs gone, what now?
Hello, I had the RAID subsystem that my root.afs volume was on crash. I'll be rebuilding that array. In the mean time is there a way to mount /afs. root.afs was the mount point for /afs, so none of the other mountpoints are valid. Without the volume online I can't even rename it. Is there a way to make the rest of my volumes available P.S. I had a RAID 5 with 8 disks in it, and apparently 2 of them crashed. Does anybody know of RAID data recovery specialists. --gabe ___ OpenAFS-info mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] win32 client 1.3.73
Hello, I have been running OpenAFS 1.2.10 on redhat 8 for sometime with linux/windows(1.2.10) clients. When I tried the 1.3.73 client on XP I experienced some difficulty. I have our entire cell (neuro.ucsd.edu) mounted to /afs, NOT /afs/neuro.ucsd.edu. I believe this is confusing the client. After a successful install, and successfully obtaining tokens, any attempt to view the mapped drive reveal only an empty directory named neuro.ucsd.edu. Is there some way to force the client to recognize my cell setup? --Gabe Castillo ___ OpenAFS-info mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-info
[OpenAFS] Trouble browsing AFS contents in XP client even though tokens are granted
Hello, I'm Running AFS client version 1.3.6390 on a windows XP Pro machine. I have a client on several other XP and linux machines and have no problems. However, The service appears to be running fine, and allows me to get a token. the root volume is mounted to z:. However, it won't let me see or manipulate any files in the afs space. It does show that out of 2TB of space that 1TB is used. Also, when in explorer if I right-click and choose AFS-Access Control List, the error that I get "unable to read access-list for directory Z: Volume name or ID not recognized" in a window titled afs_shl_ext Has anyone experienced similar problems, or have potential solutions Thanks, Gabe Castillo UC San Diego Neuroimaging Group Laboratory of Biological Dynamics and Theoretical Medicine [EMAIL PROTECTED] Do you Yahoo!?Yahoo! Domains - Claim yours for only $14.70/year
