Am 29.11.2012 um 17:53 schrieb Dan Ryson: > Hello! > > It seems my long run of good fortune has taken a turn for the worse. After > years of rock-solid AstLinux reliability, I've run into some trouble. I'm > hopeful that your expertise can help me get back to better times. > > Sorry... This is a bit of a ramble. But I'd like to provide a detailed > view of what we are seeing before soliciting your thoughts. > > Six hours after putting new hardware into service, I was INUNDATED with file > system related error messages in the log. I've pasted a snippet below, > which repeated continuously every few seconds for several hours until I shut > it down. > > This is the second time I've encountered this symptom in just a few days. > Both events were on virtually on the same hardware, a brand new NF99FL-525. > The first time this symptom appeared, I was temporarily using a USB thumb > drive in lieu of the recommended Emphase Industrial SATA module. Because I > suspected both the USB thumb drive and my twist on the official installation > procedure, I later installed the proper SATA module and asked our group for > advice regarding the general idea of copying /oldroot/mnt/asturw from one > platform to another. (Thanks again for the ideas and comments.) > Regardless, the symptom has repeated. > > Here is a more detailed version of the installation procedure that I used: > > 1.) Physdiskwrite.exe under Windows to write the SATA image > 2.) Boot from SATA on NF99FL-525 > 3.) Use web interface to create a new /dev/sda2, install sounds, etc. > 3.) Use a Linux version of FileZilla to download /oldroot/mnt/asturw from > the old Soekris net5501 > 4.) Use the same Filezilla to upload /oldroot/mnt/asturw to the new > NF99FL-525 > 5.) Reboot > 6.) Test > 7.) Put in service > > The first time around (booting from USB last Friday), the unit worked fine > for ~36 hours and then starting throwing I/O errors. When I realized this > was happening, I remotely requested a reboot. The machine shut down but > didn't come back up until I arrived on-site and cycled the power. > > This time (booting from SATA flash yesterday evening), it ran only ~6 hours > before errors appeared. Before rebooting, I downloaded the messages file > and noticed these symptoms: When I called into my voice mail, app_voicemail > generated this WARNING: "Unable to stream password file". When dialing > into our conference bridge, the dial plan tried but failed to Playback sound > file "conf-thereare". App_playback WARNED, "ast_streamfile failed for > conf-thereare." > > Upon a power cycle and reboot this morning, I noticed "fsck detected errors > on sta2" which it apparently corrected. (It went by in a blur...) After > taking the unit out of service, it's been running without I/O errors for a > couple of hours. However, it's really not really doing anything either - > except bitterly complaining about not having a network route to phones or > providers. > > Here's the log snippet. Imagine many MB of this, repeated over and over > again. Roughly 95 percent of the lines that specify a sector number, such > as "I/O error, dev sda, sector 524160" show either sector 524160 or 3145616. > Every now and again, a few other sector numbers also appear. > > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Unhandled error code > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Result: > hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] CDB: Read(10): 28 00 > 00 2f ff 90 00 00 08 00 > Nov 29 06:31:13 sip user.err kernel: end_request: I/O error, dev sda, sector > 3145616 > Nov 29 06:31:13 sip user.err kernel: EXT2-fs (sda2): previous I/O error to > superblock detected > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Unhandled error code > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Result: > hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] CDB: Write(10): 2a > 00 00 07 ff 80 00 00 08 00 > Nov 29 06:31:13 sip user.err kernel: end_request: I/O error, dev sda, sector > 524160 > Nov 29 06:31:13 sip user.err kernel: Buffer I/O error on device sda2, > logical block 0 > Nov 29 06:31:13 sip user.warn kernel: lost page write due to I/O error on > sda2 > Nov 29 06:31:13 sip user.crit kernel: EXT2-fs (sda2): error: ext2_get_inode: > unable to read inode block - inode=81936, block=327682 > Nov 29 06:31:13 sip user.err kernel: EXT2-fs (sda2): previous I/O error to > superblock detected > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Unhandled error code > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] Result: > hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK > Nov 29 06:31:13 sip user.info kernel: sd 1:0:0:0: [sda] CDB: Write(10): 2a > 00 00 07 ff 80 00 00 08 00 > Nov 29 06:31:13 sip user.err kernel: end_request: I/O error, dev sda, sector > 524160 > Nov 29 06:31:13 sip user.err kernel: Buffer I/O error on device sda2, > logical block 0 > Nov 29 06:31:13 sip user.warn kernel: lost page write due to I/O error on > sda2 > Nov 29 06:31:13 sip user.crit kernel: EXT2-fs (sda2): error: ext2_fsync: > detected IO error when writing metadata buffers > > While this really smells like a hardware problem, it seems odd that I could > have both a bad USB and bad SATA flash module. Where might I be going > wrong? Other than trying the serial version (which I'll probably do this > afternoon), what might you do in a situation like this? > > Thanks for any insight, > > Dan > > -----Original Message----- > From: Lonnie Abelbeck [mailto:li...@lonnie.abelbeck.com] > Sent: Wednesday, November 28, 2012 3:31 PM > To: AstLinux Users Mailing List > Subject: Re: [Astlinux-users] Backup => Restore to Different Platform? > > Hi Dan, > > In addition to the "70-persistent-net.rules" file as Michael pointed out, > you may (or many not) have a "/etc/rc.modules" file, you will need the > "e1000e" line uncommented for your NF99FL-525. > > Personally I prefer the geni586-serial version, but the geni586 (VGA) will > work fine if you prefer a video console. > > Lonnie > > > On Nov 28, 2012, at 8:55 AM, Dan Ryson wrote: > >> All, >> >> Following David's excellent comment about backups by rsync (pasted below), > I've been keeping copies of /oldroot/mnt/asturw. It's a simple, painless > backup procedure. >> >> I'm now replacing a production Soekris net5501 with a shiny new Jetway > NF99FL-525. Presuming I start by installing the Generic i586 (VGA) image on > the Jetway, is it permissible to simply use ssh to copy the Soekris' asturw > file tree to the new machine? If so, this would be a brain-dead simple way > to migrate to the different hardware. Or are there differences between the > two asturws that will cause pain and agony? >> >> Thanks for any thoughts, recommendations, or advice. >> >> Dan >> >> From: David Kerr [mailto:da...@kerr.net] >> Sent: Tuesday, November 20, 2012 12:13 PM >> To: AstLinux Users Mailing List >> Subject: Re: [Astlinux-users] Mixmonitor storage options >> >> I actually do a pull from a server (actually a ReadyNAS box) for my > AstLinux backup... rsync over ssh with certificates so no userid/password. > The remote server has an hourly cron job that runs the rsync.... >> >> #!/bin/bash >> rsync -avz -e "ssh -p<your-ssh-port>" > root@<your-pbx-url>:/oldroot/mnt/asturw /c/home/david/PBXbackup >> >> This way there is no dependency on anything running on the Astlinux box. > Something similar would work to pull off call recordings for safe storage. >> >> David
Hi Dan, it is possible, that there are other "hardware-related" files in ASTURW like e.g. "/etc/blkid.tab*" I personally wouldn't clone a complete device. Maybe start with a basic backup of the old box WITHOUT including ASTURW. Michael http://www.mksolutions.info
smime.p7s
Description: S/MIME cryptographic signature
------------------------------------------------------------------------------ Keep yourself connected to Go Parallel: VERIFY Test and improve your parallel project with help from experts and peers. http://goparallel.sourceforge.net
_______________________________________________ Astlinux-users mailing list Astlinux-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/astlinux-users Donations to support AstLinux are graciously accepted via PayPal to pay...@krisk.org.
