Re: amanda 3.3.3 sellfcheck request failed
Work-around. I was able to get amanda to work on the client using "bsd" protocal, rather than bsdtcp. This is not optimal, but is an improvement over using the 2.6.1 client. Which to tell the truth worked pretty darn well, with the one nagging issue being the fact that the client daemons would often hang after job completion, causing problems dumping the 'pending' DLE the next day. We'll see how the client behaves this evening. On Thu, May 16, 2013 at 03:11:33PM -0400, Brian Cuttler wrote: > On Thu, May 16, 2013 at 03:02:53PM -0400, Jean-Louis Martineau wrote: > > On 05/16/2013 02:53 PM, Brian Cuttler wrote: > > >/tmp/amanda/amandad - which was created by the manual run > > >of amandad, no new files. > > > > If telnet do not create a new debug file it is because amandad is not > > executed, which means SMF is misconfigured. > > I can't help with SMF. > > OS specific issue, I can appreciate that. Thank you. > > Does anyone have a clue as to how I've misconfigured SMF on Solaris 10? > > My amanda config on Finsen, which is a solaris 10/x86 amanda server > that has only itself as a client, has the following entry in inetd.conf > > amanda stream tcp nowait amanda /usr/local/libexec/amanda/amandad >amandad -auth=bsdtcp amdump > > This is identical to the entry in inetd.conf on Grifserv, the new > amanda client I'm trying to upgrade for amanda server Curie. > > service output information is identical on the two machines. > > # svcs -l svc:/network/amanda/tcp:default > fmri svc:/network/amanda/tcp:default > name amanda > enabled true > stateonline > next_state none > state_time Thu May 16 14:31:24 2013 > restartersvc:/network/inetd:default > contract_id > > The manifest services give every indication of being identical. > > [finsen]: /var/svc/manifest/network > ls -l *amanda* > -rw-r--r-- 1 root root 2320 Feb 9 01:05 amanda-tcp.xml > -rw-r--r-- 1 root root 2485 Feb 9 01:05 amanda-udp.xml > -rw-r--r-- 1 root root 2292 Feb 27 2009 amandaidx-tcp.xml > > # ls -l *amanda* > -rw-r--r-- 1 root root2320 May 16 14:31 amanda-tcp.xml > > # wc amanda-tcp.xml > 82 2272320 amanda-tcp.xml > > # cksum amanda-tcp.xml > 4135061186 2320amanda-tcp.xml > > ... I'm going to have another look at the dumptype definitions. > > > > > > > > --- >Brian R Cuttler brian.cutt...@wadsworth.org >Computer Systems Support(v) 518 486-1697 >Wadsworth Center(f) 518 473-6384 >NYS Department of HealthHelp Desk 518 473-0773 > --- Brian R Cuttler brian.cutt...@wadsworth.org Computer Systems Support(v) 518 486-1697 Wadsworth Center(f) 518 473-6384 NYS Department of HealthHelp Desk 518 473-0773
Re: amanda 3.3.3 sellfcheck request failed
On Thu, May 16, 2013 at 03:02:53PM -0400, Jean-Louis Martineau wrote: > On 05/16/2013 02:53 PM, Brian Cuttler wrote: > >/tmp/amanda/amandad - which was created by the manual run > >of amandad, no new files. > > If telnet do not create a new debug file it is because amandad is not > executed, which means SMF is misconfigured. > I can't help with SMF. OS specific issue, I can appreciate that. Thank you. Does anyone have a clue as to how I've misconfigured SMF on Solaris 10? My amanda config on Finsen, which is a solaris 10/x86 amanda server that has only itself as a client, has the following entry in inetd.conf amanda stream tcp nowait amanda /usr/local/libexec/amanda/amandad amandad -auth=bsdtcp amdump This is identical to the entry in inetd.conf on Grifserv, the new amanda client I'm trying to upgrade for amanda server Curie. service output information is identical on the two machines. # svcs -l svc:/network/amanda/tcp:default fmri svc:/network/amanda/tcp:default name amanda enabled true stateonline next_state none state_time Thu May 16 14:31:24 2013 restartersvc:/network/inetd:default contract_id The manifest services give every indication of being identical. [finsen]: /var/svc/manifest/network > ls -l *amanda* -rw-r--r-- 1 root root 2320 Feb 9 01:05 amanda-tcp.xml -rw-r--r-- 1 root root 2485 Feb 9 01:05 amanda-udp.xml -rw-r--r-- 1 root root 2292 Feb 27 2009 amandaidx-tcp.xml # ls -l *amanda* -rw-r--r-- 1 root root2320 May 16 14:31 amanda-tcp.xml # wc amanda-tcp.xml 82 2272320 amanda-tcp.xml # cksum amanda-tcp.xml 4135061186 2320amanda-tcp.xml ... I'm going to have another look at the dumptype definitions. --- Brian R Cuttler brian.cutt...@wadsworth.org Computer Systems Support(v) 518 486-1697 Wadsworth Center(f) 518 473-6384 NYS Department of HealthHelp Desk 518 473-0773
Re: amanda 3.3.3 sellfcheck request failed
On 05/16/2013 02:53 PM, Brian Cuttler wrote: /tmp/amanda/amandad - which was created by the manual run of amandad, no new files. If telnet do not create a new debug file it is because amandad is not executed, which means SMF is misconfigured. I can't help with SMF. Jean-Louis
Re: amanda 3.3.3 sellfcheck request failed
/tmp/amanda/amandad - which was created by the manual run of amandad, no new files. On Thu, May 16, 2013 at 02:50:37PM -0400, Jean-Louis Martineau wrote: > On 05/16/2013 02:46 PM, Brian Cuttler wrote: > >But I'm not seeing, perhaps, looking in the wrong directory, any > >debug files. > > In: `amgetconf build.amanda_dbgdir`/amandad > > Jean-Louis --- Brian R Cuttler brian.cutt...@wadsworth.org Computer Systems Support(v) 518 486-1697 Wadsworth Center(f) 518 473-6384 NYS Department of HealthHelp Desk 518 473-0773
Re: amanda 3.3.3 sellfcheck request failed
On 05/16/2013 02:46 PM, Brian Cuttler wrote: But I'm not seeing, perhaps, looking in the wrong directory, any debug files. In: `amgetconf build.amanda_dbgdir`/amandad Jean-Louis
Re: amanda 3.3.3 sellfcheck request failed
Jean-Louis, On Thu, May 16, 2013 at 02:40:28PM -0400, Jean-Louis Martineau wrote: > Brian, > > Do a: telnet grifserv.wadsworth.org amanda > create an amandad debug file on grifserv? > telnet grifserv amanda Trying 10.49.66.6... Connected to grifserv.wadsworth.org. Escape character is '^]'. Connection to grifserv.wadsworth.org closed by foreign host. produces # snoop host curie Using device bge0 (promiscuous mode) curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=59114 Syn Seq=532791431 Len=0 Win=49640 Options= grifserv.wadsworth.org -> curie.wadsworth.org TCP D=59114 S=10080 Syn Ack=532791432 Seq=2245896368 Len=0 Win=49640 Options= curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=59114 Ack=2245896369 Seq=532791432 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=59114 S=10080 Fin Ack=532791432 Seq=2245896369 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=59114 Ack=2245896370 Seq=532791432 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=59114 Fin Ack=2245896370 Seq=532791432 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=59114 S=10080 Ack=532791433 Seq=2245896370 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=51483 Push Ack=97669462 Seq=1337214714 Len=396 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=33469 Syn Seq=2993683995 Len=0 Win=49640 Options= curie.wadsworth.org -> grifserv.wadsworth.org TCP D=33469 S=9997 Syn Ack=2993683996 Seq=543998241 Len=0 Win=49640 Options= grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=33469 Ack=543998242 Seq=2993683996 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=33469 Fin Ack=543998242 Seq=2993683996 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=33469 S=9997 Ack=2993683997 Seq=543998242 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=33469 S=9997 Fin Ack=2993683997 Seq=543998242 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=33469 Ack=543998243 Seq=2993683997 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=51483 S=9997 Ack=1337215110 Seq=97669462 Len=0 Win=48419 But I'm not seeing, perhaps, looking in the wrong directory, any debug files. > > Jean-Louis > > On 05/16/2013 02:06 PM, Brian Cuttler wrote: > >Hello amanda users, > > > >And the great wheel turns again... I've been here before, but > >apparently fell back to 2.6.1 and upd from a failed 3.3.0 install. > > > > > >I am running amanda 3.1.2 on Solaris 10/x86 as a server and > >am in the process of upgrading the amanda client on solaris 10/Sparc > >from 2.6.1 to 3.3.3. > > > >Built seems to have gone fine. > > > >Problme is the amcheck is failing. > > > >I have updated the xinetd.conf on the client, it now reads. > > > >amanda stream tcp nowait amanda /usr/local/libexec/amanda/amandad > > amandad -auth=bsdtcp amdump > > > >I disabled the prior instance of amanda (udp) in SMF services and > >removed the config file both from the SMF manifest directory and from > >the compiled database and then imported the new inetd config. > > > >I updated the disklist, rather than using > > > > > >define dumptype zfs-snapshot { > > index > > program "APPLICATION" > > application "app_amgtar" > > script "script_zfs_snapshot" > ># auth "bsdtcp" > > estimate server > >} > > > >We use > > > >define dumptype zfs-snapshot-bsdtcp { > > index > > program "APPLICATION" > > application "app_amgtar" > > script "script_zfs_snapshot" > > auth "bsdtcp" > > estimate server > >} > > > > > >I have a note that says the auto param should be in the script and > >not the dumptype, but that produces an immediate error when I invoke > >amcheck. > > > > > >I have updated exactly one DLE in the disklist for this client > >and I have commented out all other references to this host in > >the disklist, this is to prevent any confusion as to which protocal > >should be used. > > > >I am not seeing any errors in /var/adm/mesasges, nor are we producing > >any debug files in the /tmp/amanda directory. > > > >Amcheck on the server does this. > > > >>amcheck -c curie grifserv > >Amanda Backup Client Hosts Check > > > >WARNING: grifserv.wadsworth.org: selfcheck request failed: EOF on read > >from grifserv.wadsworth.org > >Client check: 1 host checked in 0.077 seconds. 1 problem found. > > > >(brought to you by Amanda 3.1.2) > > > >Snoop between the server and the client produces this > > > >curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Syn > >Seq=3583579017 Len=0 Win=49640 Options= >0,nop,nop,sackOK> > >grifserv.wadsworth.org -> curie.wadsworth.org TCP D=516 S=10080 Syn > >Ack=3583579018 Seq=144367203 Len=0 Win=49640 Options= >0,nop,nop,sackOK> > >curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 > >Ack=144367204 Seq=3583579018 Len=0 Win=49640 > >
Re: amanda 3.3.3 sellfcheck request failed
Brian, Do a: telnet grifserv.wadsworth.org amanda create an amandad debug file on grifserv? Jean-Louis On 05/16/2013 02:06 PM, Brian Cuttler wrote: Hello amanda users, And the great wheel turns again... I've been here before, but apparently fell back to 2.6.1 and upd from a failed 3.3.0 install. I am running amanda 3.1.2 on Solaris 10/x86 as a server and am in the process of upgrading the amanda client on solaris 10/Sparc from 2.6.1 to 3.3.3. Built seems to have gone fine. Problme is the amcheck is failing. I have updated the xinetd.conf on the client, it now reads. amanda stream tcp nowait amanda /usr/local/libexec/amanda/amandad amandad -auth=bsdtcp amdump I disabled the prior instance of amanda (udp) in SMF services and removed the config file both from the SMF manifest directory and from the compiled database and then imported the new inetd config. I updated the disklist, rather than using define dumptype zfs-snapshot { index program "APPLICATION" application "app_amgtar" script "script_zfs_snapshot" # auth "bsdtcp" estimate server } We use define dumptype zfs-snapshot-bsdtcp { index program "APPLICATION" application "app_amgtar" script "script_zfs_snapshot" auth "bsdtcp" estimate server } I have a note that says the auto param should be in the script and not the dumptype, but that produces an immediate error when I invoke amcheck. I have updated exactly one DLE in the disklist for this client and I have commented out all other references to this host in the disklist, this is to prevent any confusion as to which protocal should be used. I am not seeing any errors in /var/adm/mesasges, nor are we producing any debug files in the /tmp/amanda directory. Amcheck on the server does this. amcheck -c curie grifserv Amanda Backup Client Hosts Check WARNING: grifserv.wadsworth.org: selfcheck request failed: EOF on read from grifserv.wadsworth.org Client check: 1 host checked in 0.077 seconds. 1 problem found. (brought to you by Amanda 3.1.2) Snoop between the server and the client produces this curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Syn Seq=3583579017 Len=0 Win=49640 Options= grifserv.wadsworth.org -> curie.wadsworth.org TCP D=516 S=10080 Syn Ack=3583579018 Seq=144367203 Len=0 Win=49640 Options= curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Ack=144367204 Seq=3583579018 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Push Ack=144367204 Seq=3583579018 Len=89 Win=64240 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=516 S=10080 Ack=3583579107 Seq=144367204 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=516 S=10080 Fin Ack=3583579107 Seq=144367204 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Ack=144367205 Seq=3583579107 Len=0 Win=64240 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=10080 S=516 Fin Ack=144367205 Seq=3583579107 Len=0 Win=64240 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=516 S=10080 Ack=3583579108 Seq=144367205 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=51483 Push Ack=97669462 Seq=1337171445 Len=396 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=65528 Syn Seq=3928298379 Len=0 Win=49640 Options= curie.wadsworth.org -> grifserv.wadsworth.org TCP D=65528 S=9997 Syn Ack=3928298380 Seq=3586269716 Len=0 Win=49640 Options= grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=65528 Ack=3586269717 Seq=3928298380 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=65528 Fin Ack=3586269717 Seq=3928298380 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=65528 S=9997 Ack=3928298381 Seq=3586269717 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=65528 S=9997 Fin Ack=3928298381 Seq=3586269717 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=65528 Ack=3586269718 Seq=3928298381 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=51483 S=9997 Ack=1337171841 Seq=97669462 Len=0 Win=48419 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=51483 Push Ack=97669462 Seq=1337171841 Len=396 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=32768 Syn Seq=342358183 Len=0 Win=49640 Options= curie.wadsworth.org -> grifserv.wadsworth.org TCP D=32768 S=9997 Syn Ack=342358184 Seq=3599432777 Len=0 Win=49640 Options= grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=32768 Ack=3599432778 Seq=342358184 Len=0 Win=49640 grifserv.wadsworth.org -> curie.wadsworth.org TCP D=9997 S=32768 Fin Ack=3599432778 Seq=342358184 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=32768 S=9997 Ack=342358185 Seq=3599432778 Len=0 Win=49640 curie.wadsworth.org -> grifserv.wadsworth.org TCP D=32768 S=9997 Fin Ack=342358185 Seq=3599432778 Len=0 Win=49640 grifserv.wadsworth.