Ethan Erchinger wrote:
> Hello,
> We have a backup strategy that involves mapping LUNs between a given 
> pair of hosts, and copying data from one of the LUNs (src) and another 
> LUN (dest).  The src LUNs sit a SAN device, sometimes multiple devices 
> (zpool mirror).  The src LUN is running a MySQL database and typically 
> will be running for weeks without issue.
>
> When we start the backup sequence, we map a previously unmapped LUN to 
> the DB host and issue the following commands:
>
> root# cfgadm -al
> (sleep 10)
> root# luxadm probe
> (sleep 10)
> root# zpool import <pool_name>
>
> After importing we'll perform some minor IO on the dest LUN, such as 
> adding a symlink, removing some old configuration files.  Then we'll 
> start an ibbackup of that database from the src LUN to the dest LUN, 
> and things go bad.
>
> It's not completely consistent, but sometimes the DB host will crash, 
> sometimes we'll get chksum/read/write errors on the src LUN.  Looking 
> at dmesg (when the host doesn't crash), we see the LUNs paths all 
> disappear and then reappear usually around 20 seconds later.  Example 
> output below.  Each LUN has 2 paths out of the DB host and 4 paths on 
> each storage device, across two separate SANs.
>
> Usually the host will crash when not running with a zpool mirror, 
> which apparently in Sol10u4, it's expected behavior.
>
> These hosts are x86_64 servers, running Sol10u4, unpatched.  They use 
> qlogic qla2342 HBAs, and the stock qlc driver.  They are using MPXIO, 
> from what I can tell.
>
> If anyone has any tips on troubleshooting, or knows of things we are 
> doing wrong, help would be appreciated.


Hey Ethan,

 Gotta tell you, this is a pretty gutsy and interesting backup solution. :)

 While I'll give you props for coming up with something innovative I'd 
strongly recommend against this type of solution.  Manipulation like 
this really is flirting with a panic.  Adding and removing devices too 
frequently on FC is in my experience dangerous in production environments.

 If you want to stick with a solution on FC I'd recommend looking at AVS 
Instant Image and/or Remote Mirror 
(http://www.opensolaris.org/os/project/avs/).  That'll give you a proper 
replication solution.

 Since your running ZFS, I _highly_ recommend snapshotting the 
filesystem containing MySQL data and then using zfs send/recv to move 
the data to a remote host. 

  Since you paid good money for ibbackup, I'm going to assume you like 
it and have a production database that you want to be gentle too.  I'd 
consider using NFS as a backup destination or providing a backup 
destintation LUN to the server and just leaving it, perhaps using UFS so 
that you could mount it elsewhere easily without importing/exporting 
LUNs on the host.

 My philosophy with FC on Solaris in a production environment is to set 
it and step away.  If you can utilize ethernet, particularly with ZFS 
send/recv I think you'll be happier and avoid the cost of ibbackup.


 I didn't really answer your question, but hopefully there are some 
options here that might be more appealing.  If you want additional 
detail on any of them please let me know.

 benr.
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to