Can I reboot during a metadevice resync?


 
Thread Tools Search this Thread
Top Forums UNIX for Dummies Questions & Answers Can I reboot during a metadevice resync?
# 1  
Old 08-29-2011
Question Can I reboot during a metadevice resync?

In my infinite lack of wisdom, I brought a Solaris 10(Sparc, 64 bit ) system down to init 1, detached a metadrive so that I could run format->analyze->refresh on it, and then reinit'd and reattached it .

It started syncing the drive as it's a submirror of a mirrored drive, which happens contain the filesystem for a production application.

Ideally, I would have rebooted the system to get the application back up before re-attaching and resyncing, but I did not.

Would it by safe to reboot the system while the submirror is rsync'ing? Or do I have to wait until it is done?
# 2  
Old 08-29-2011
As far as I know it doesn't matter , If I were in your shoe I'll Just wait until resync finishes.
# 3  
Old 08-30-2011
Quote:
Would it by safe to reboot the system while the submirror is rsync'ing? Or do I have to wait until it is done?
Probably not a good idea.
# 4  
Old 08-30-2011
Hammer & Screwdriver

I waited, the question was more for a "next time this happens..." situation.
Thread closed.
# 5  
Old 08-30-2011
yes you can but not recommended.

you don't need to reboot to bring it back to runlevel 3. just run init again
# 6  
Old 08-30-2011
Hmm what happens if you ctrl-C a copy, cp operation( just to use an analogy)? it is probably the same when SDS/SVM runs a sync, block by block copy from source to destination (one sub-mirror to another sub-mirror), assuming RAID 1 here.
So if you stop the copy, the source is not getting affected, so you are left with the good data on the source side. Well the destination will have half cooked data which is not going to be of any use.

And when you run boot on the ok OBP prompt, I don't think OBP has any knowledge or understanding the SDS/SVM disk, so it will read the boot block off the physical drive path to get a start. The md driver is loaded once the kernel loads them in the boot process. So what happens if the first boot disk is the one that has the half cooked data from above, it ain't going to boot Smilie. You still would be able to boot from the other good disk.

So, there was the technical how and why Smilie, which is always different from the decisions you have to make based on the risk you want to take...in the end the question remains, is it worth doing to go through the trouble Smilie, probably not. So let it finish what it is doing. Sometimes the benefits of waiting outweigh the risk of being hasty.

Correct me if I am wrong or stated a fact wrong. I would be happy someone reviewed this and provided some commentary on the above,
# 7  
Old 08-30-2011
Quote:
Originally Posted by dude2cool
And when you run boot on the ok OBP prompt, I don't think OBP has any knowledge or understanding the SDS/SVM disk, so it will read the boot block off the physical drive path to get a start. The md driver is loaded once the kernel loads them in the boot process. So what happens if the first boot disk is the one that has the half cooked data from above, it ain't going to boot Smilie.
Check the documentation on the Volume Read and Write polices. You won't have "half cooked" data if setup properly.

The OP did not specify that the drive is a boot drive. Did I miss something? Even if it was a boot drive it would still work if setup properly. There was no technical reason to bring the system to runlevel 1. The power of mirroring or other types of RAID is that it can be done on a live system. If its a big disk the resync will be slow regardless of the runlevel.

That being said I would not reboot it but the OP can run init 3 to bring it back to multi-user level without a reboot.

my .02

Last edited by frank_rizzo; 08-30-2011 at 10:46 PM.. Reason: added content
 
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. Shell Programming and Scripting

Trying to make a resync script

Good Morning, I'm trying to automate the resync process after a mirror replacement on Solaris. To avoid resyncing all the drives at once, I'm trying to use metastat for each drive to determine if the drive is still resyncing or if it is done (Okay). The following test lines seem to always show... (4 Replies)
Discussion started by: Stellaman1977
4 Replies

2. Solaris

Reboot causes disks in Resync State

Dear Team, This time i am facing some new problems which is beyond my thinking. I need some expert advice. We are having 4 Servers ( 2 nos Sun SPARC Enterprise T5220 & 2 nos of SF e2900 Servers ). Both the T5220 Servers are Termed as Node A and Node B . The same things are followed with... (1 Reply)
Discussion started by: sudhansu
1 Replies

3. AIX

Resync data on File system

Hi All, I have two mount points have the same data with little changes between them /appabc1 /appabc2 Both of them have the same data, there is some little changes on the data between them I want /appabc2 to has the same data of /appabc1 exactly including to those little changes... (6 Replies)
Discussion started by: Mr.AIX
6 Replies

4. Solaris

metadevice submirrors 0 and 1 not consistent

At a new job, I've inherited a handful of Solaris 10 SunFire V240 servers. I noticed that the metadevices are not consistent, in a way that I've not seen before.... internal mirrored disks: metadevice d60 has c1t0d0s6 as submirror 0 (d61) and c1t1d0s6 as submirror 1 (d62) metadevice... (1 Reply)
Discussion started by: Max Power
1 Replies

5. Solaris

metadevice problem

hi all i am very new to solaris...i am trying configure raid 0 in my ultrascsi box which is having 6 sun36GB hard disks. I have created 4 state database replicas in the following slices /dev/dsk/c1t1d0s7 /dev/dsk/c1t2d0s7 /dev/dsk/c1t3d0s7 and /dev/dsk/c1t4d0s7. i could not use any other... (7 Replies)
Discussion started by: kingston
7 Replies

6. Solaris

increase metadevice size

Hi, I am having two metadevices d50 and d100 which are used to created soft partitions as and when required. d50 and d100 are metadevices formed on different disks. d50 -- disks 0 & 1 d100 -- disks 2 & 3 I have a soft partition d70 os 50 GB on d50. Now there is no free space on d50. ... (1 Reply)
Discussion started by: sag71155
1 Replies

7. Shell Programming and Scripting

trailing slash - resync command

Hi All, i am aware that there is a difference between: 1. rsync -n -av /tmp . 2. rsync -n -av /tmp/ . I would like to do the first option. But if i use a variable (rsync -av $log .), the command behaves like a second option (with the trailing slash) Is there any way to use a... (2 Replies)
Discussion started by: c00kie88
2 Replies

8. AIX

How long does AIX resync the time with another server

Hi, I have NTP configured: vi /etc/ntp.conf broadcastclient server 128.127.1.3 driftfile /etc/ntp.drift tracefile /etc/ntp.trace # xntpdc xntpdc> sysinfo system peer: 128.127.1.3 system peer mode: client leap indicator: 00 stratum: 12 precision: ... (0 Replies)
Discussion started by: victorcheung
0 Replies

9. Linux

HELP PLEASE about Resync and sync interval in linux

Can anyone could tell me what is the meaning of this problem: Last successful check resync is greater than min sync interval (1195785433 > 7200) And what can do to solve this problem. Thank you..Please HElp me! (4 Replies)
Discussion started by: o_m_g
4 Replies

10. UNIX for Dummies Questions & Answers

raidctl and resync when reboot

I am using raidctl on a v440 disk and noticed it resyncs after every boot, which takes about 30 minutes because of the size of the partition. I am concerned with what happens during the resync if "writes" happen to the disk before it is complete? Any info would be helpful. Thanks (0 Replies)
Discussion started by: csgonan
0 Replies
Login or Register to Ask a Question