I have followed the Wiki instructions on how to upgrade a drive with a bigger one and it took forever. I had a 500GB drive and replaced it with a 1.5TB. The amount of data that was on the original drive was about 250GB and the replacement took about 5 days.

I want to replace another 500GB drive with a new 1.5TB and I was thinking will it be faster to move the data on the 500GB drive, remove it from the array, then install the 1.5TB and add it to the array.

 

Will that take less then 5 days? is there any risk on doing that?

 

Thanks,

Tal

What unraid version are you using?

 

How many drives do you have? What are there sizes?

 

After copying the data there will be need to run the parity anyway.

 

There could be a speed (read/writing) problem during parity and regular reads/writes that causes the long update.

 

UnRaid 4.5

I have total of 6 drives. 3x 1.5TB, 1x 1TB, and 2x500GB when I did the first upgrade.

The system was ideal for about 80% of the time.

Read/Write speed tests that I did was between 50MBs to 70MBs if I remember correctly.

What's your system specs (cpu, motherboard, memory, sata or ide, if sata using AHCI)?

 

There's no reason it should take 5 days to do anything unless you have 16 drives hanging off the PCI-Bus or all your other drives are having severe issues.

My system is a pretty powerfull machine for an UnRaid system.

 

My system specs are:

CPU: Intel® Core2 Duo CPU     E7400  @ 2.80GHz

Mem:CORSAIR 2GB (2 x 1GB) 240-Pin DDR3 SDRAM DDR3 1066 (PC3 8500) Dual Channel Kit Desktop Memory Model TWIN3X2048-1066C7

MB: SUPERMICRO MBD-C2SEE-O LGA 775 Intel G43 ATX Intel Motherboard

Controllers: 2x Adaptec 2240900-R PCI Express 4-lane 2.5 Gb/s Low Profile Ready SATA 1430SA Kit

HD Cage: SUPERMICRO CSE-M35T-1B Black 5 Bay Hot-Swapable SATA HDD Enclosure

 

I do not remember if I use AHCI but all 6 drives are connected to the Adaptec SATA controllers and nothing is connected to the MB.    

That's very odd indeed as nothing should be the cause of any bottlenecks.

 

If you haven't rebooted then if you can, post a syslog to see if there's device resets or other nasty things going on that you might not be aware of.

 

For what it's worth, I'd run all of those drives off the MB SATA ports, but that's just my personal preference to limit add-on cards unless absolutely needed.

UnRaid 4.5

I have total of 6 drives. 3x 1.5TB, 1x 1TB, and 2x500GB when I did the first upgrade.

 

Are any of those WD 'Advanced Format' drives (EARS models)?

My 1.5 TB are Seagate drives with the updated firmware - ST31500341AS

The 1Tb is a Samsung drive - SAMSUNG HD103UJ

The new 1.5TB also is a Samsung drive - SAMSUNG HD154UI

The one that I have replaced is a WD 500Gb but it is not the model you have mentioned.

 

Can you break down the 5-days for us?

 

Which steps did you run in what order and how long did each step take?

Day 1:

Shut down the arry.

Replaced old 500GB drive with 1.5TB drive

Powered up the machine

Started the Array

Let the Array rebuild

Day 2 ,3, 4 and 5:

Array rebuilding with minimum use of the whole array. I did look at the array while it was rebuilding and all drives were on. I found that odd since I have my drives set to spin down on 15min of inactivity.

 

I have another 500GB drive that I want to replace with a new 1.5TB and if it takes 5 days, that is fine. If it should take way less then that, I would like to know what options I have. I thought about moving all the data on the 500GB drive to another drive, doing a parity check, deleting the 500GB drive from the array, and then adding the new 1.5TB drive to the array. I am sure this will take less then 5 days.

 

Tal

I have followed the Wiki instructions on how to upgrade a drive with a bigger one and it took forever. I had a 500GB drive and replaced it with a 1.5TB. The amount of data that was on the original drive was about 250GB and the replacement took about 5 days.

I want to replace another 500GB drive with a new 1.5TB and I was thinking will it be faster to move the data on the 500GB drive, remove it from the array, then install the 1.5TB and add it to the array.

 

Will that take less then 5 days? is there any risk on doing that?

 

Thanks,

Tal

 

FYI, i recently upgrade the parity disk and one data disk in my unRAID to 2TB, whole process took less than 24 hours. my unRAID has 14 disks.

 

Any given time, if there is an issue, posted your syslog is the first step

 

 

Here is my syslog. If I need to post one right after I restart the machine, Please let me know.

 

Sep  5 22:44:14 Tower emhttp: shcmd (20): echo '# Generated mover schedule:' >>/var/spool/cron/crontabs/root

Sep  5 22:44:14 Tower emhttp: shcmd (21): echo '40 3 * * * /usr/local/sbin/mover 2>&1 | logger' >>/var/spool/cron/crontabs/root

Sep  5 22:44:14 Tower emhttp: shcmd (22): crontab /var/spool/cron/crontabs/root

Sep  5 22:44:15 Tower emhttp: shcmd (23): rm /etc/samba/smb-shares.conf >/dev/null 2>&1

Sep  5 22:44:15 Tower emhttp: _shcmd: shcmd (23): exit status: 1

Sep  5 22:44:15 Tower emhttp: shcmd (24): cp /etc/exports- /etc/exports

Sep  5 22:44:15 Tower emhttp: shcmd (25): mkdir /mnt/user

Sep  5 22:44:15 Tower emhttp: shcmd (26): /usr/local/sbin/shfs /mnt/user  -o noatime,big_writes,allow_other,default_permissions

Sep  5 22:44:16 Tower emhttp: shcmd (27): killall -HUP smbd

Sep  5 22:44:16 Tower emhttp: shcmd (28): /etc/rc.d/rc.nfsd restart | logger

Sep  5 22:45:00 Tower cache_dirs: ==============================================

Sep  5 22:45:00 Tower cache_dirs: command-args=-w -m 1 -M 10 -d 9999 -e Backup -e CuddleSleeve -a -noleaf

Sep  5 22:45:00 Tower cache_dirs: vfs_cache_pressure=10

Sep  5 22:45:00 Tower cache_dirs: max_seconds=10, min_seconds=1

Sep  5 22:45:00 Tower cache_dirs: max_depth=9999

Sep  5 22:45:00 Tower cache_dirs: command=find -noleaf

Sep  5 22:45:00 Tower cache_dirs: version=1.6.4

Sep  5 22:45:00 Tower cache_dirs: ---------- caching directories ---------------

Sep  5 22:45:00 Tower cache_dirs: DS

Sep  5 22:45:00 Tower cache_dirs: Disks

Sep  5 22:45:00 Tower cache_dirs: Home Movies

Sep  5 22:45:00 Tower cache_dirs: ImagesByName

Sep  5 22:45:00 Tower cache_dirs: Movies

Sep  5 22:45:00 Tower cache_dirs: Music

Sep  5 22:45:00 Tower cache_dirs: PSP Games

Sep  5 22:45:00 Tower cache_dirs: Pictures

Sep  5 22:45:00 Tower cache_dirs: Stuff

Sep  5 22:45:00 Tower cache_dirs: TV Shows

Sep  5 22:45:00 Tower cache_dirs: Trailers

Sep  5 22:45:00 Tower cache_dirs: ----------------------------------------------

Sep  5 22:45:00 Tower cache_dirs: cache_dirs process ID 1853 started, To terminate it, type: cache_dirs -q

Sep  6 02:06:04 Tower kernel: mdcmd (1255): spindown 5

Sep  6 05:35:46 Tower kernel: mdcmd (2504): spindown 3

Sep  6 07:55:07 Tower kernel: md: sync done. time=33051sec rate=44329K/sec

Sep  6 07:55:07 Tower kernel: md: recovery thread sync completion status: 0

Sep  6 08:25:13 Tower kernel: mdcmd (3514): spindown 0

Sep  6 08:25:13 Tower kernel: mdcmd (3515): spindown 1

Sep  6 08:25:13 Tower kernel: mdcmd (3516): spindown 2

Sep  6 08:25:13 Tower kernel: mdcmd (3517): spindown 4

Sep  6 12:15:09 Tower kernel: mdcmd (4899): spindown 4

Sep  6 12:15:20 Tower kernel: mdcmd (4901): spindown 0

Sep  6 12:15:21 Tower kernel: mdcmd (4902): spindown 1

Sep  6 21:51:34 Tower kernel: mdcmd (8336): spindown 5

Sep  7 16:29:51 Tower kernel: mdcmd (14999): spindown 1

Sep  7 16:29:51 Tower kernel: mdcmd (15000): spindown 5

Sep  7 19:22:09 Tower kernel: mdcmd (16027): spindown 0

Sep  7 20:03:15 Tower kernel: mdcmd (16271): spindown 5

Sep  7 20:05:36 Tower kernel: mdcmd (16286): spindown 1

Sep  7 20:05:56 Tower kernel: mdcmd (16289): spindown 0

Sep  7 20:05:56 Tower kernel: mdcmd (16290): spindown 4

Sep  7 21:16:12 Tower kernel: mdcmd (16707): spindown 3

Sep  7 21:46:31 Tower kernel: mdcmd (16888): spindown 3

Sep  7 21:46:42 Tower kernel: mdcmd (16890): spindown 2

Sep  7 21:46:52 Tower kernel: mdcmd (16892): spindown 5

Sep  7 21:49:14 Tower kernel: mdcmd (16907): spindown 1

Sep  7 21:50:54 Tower kernel: mdcmd (16918): spindown 0

Sep  7 21:50:55 Tower kernel: mdcmd (16919): spindown 4

Sep  7 23:56:09 Tower kernel: mdcmd (17666): spindown 4

Sep  8 10:29:41 Tower kernel: mdcmd (21441): spindown 0

Sep  8 11:01:45 Tower kernel: mdcmd (21633): spindown 1

Sep  8 16:02:33 Tower kernel: mdcmd (23426): spindown 4

Sep  8 21:44:13 Tower kernel: mdcmd (25462): spindown 4

Sep  8 23:59:12 Tower kernel: mdcmd (26267): spindown 2

Sep  9 04:44:59 Tower kernel: mdcmd (27969): spindown 1

Sep  9 04:52:12 Tower kernel: mdcmd (28013): spindown 0

Sep  9 04:52:12 Tower kernel: mdcmd (28014): spindown 4

Sep  9 08:31:03 Tower kernel: mdcmd (29317): spindown 0

Sep  9 08:31:03 Tower kernel: mdcmd (29318): spindown 4

Sep  9 09:08:41 Tower kernel: mdcmd (29542): spindown 0

Sep  9 09:08:41 Tower kernel: mdcmd (29543): spindown 4

Sep  9 21:26:03 Tower kernel: mdcmd (33934): spindown 2

Sep  9 21:26:24 Tower kernel: mdcmd (33937): spindown 3

Sep  9 21:26:45 Tower kernel: mdcmd (33940): spindown 5

Sep  9 21:26:55 Tower kernel: mdcmd (33942): spindown 1

Sep  9 21:27:46 Tower kernel: mdcmd (33948): spindown 0

Sep  9 22:56:33 Tower kernel: mdcmd (34478): spindown 4

Sep 10 20:14:31 Tower kernel: mdcmd (42092): spindown 0

Sep 10 20:14:31 Tower kernel: mdcmd (42093): spindown 4

Sep 10 21:21:26 Tower kernel: mdcmd (42491): spindown 0

Sep 10 21:21:26 Tower kernel: mdcmd (42492): spindown 1

Sep 10 22:16:19 Tower kernel: mdcmd (42820): spindown 4

Sep 11 11:59:45 Tower kernel: mdcmd (47726): spindown 2

Sep 11 11:59:56 Tower kernel: mdcmd (47728): spindown 5

Sep 11 12:00:16 Tower kernel: mdcmd (47731): spindown 4

Sep 11 12:02:28 Tower kernel: mdcmd (47745): spindown 0

Sep 11 12:02:28 Tower kernel: mdcmd (47746): spindown 1

Sep 11 14:25:15 Tower kernel: mdcmd (48597): spindown 0

Sep 11 14:25:15 Tower kernel: mdcmd (48598): spindown 1

Sep 11 18:23:52 Tower kernel: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen

Sep 11 18:23:52 Tower kernel: ata2.00: cmd e5/00:00:00:00:00/00:00:00:00:00/00 tag 0

Sep 11 18:23:52 Tower kernel:          res 40/00:ff:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout)

Sep 11 18:23:52 Tower kernel: ata2.00: status: { DRDY }

Sep 11 18:23:52 Tower kernel: ata2: hard resetting link

Sep 11 18:23:53 Tower kernel: ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Sep 11 18:23:53 Tower kernel: ata2.00: configured for UDMA/133

Sep 11 18:23:53 Tower kernel: ata2: EH complete

Sep 11 18:33:27 Tower kernel: mdcmd (50073): spindown 2

Sep 11 18:33:37 Tower kernel: mdcmd (50075): spindown 5

Sep 11 18:33:58 Tower kernel: mdcmd (50078): spindown 1

Sep 12 09:29:47 Tower kernel: mdcmd (55416): spindown 5

 

Here is my syslog. If I need to post one right after I restart the machine, Please let me know.

 

If you still have whole syslog, from boot to now, please post it

 

Attach it as a zip file.  Do not try to post it as text.  It will be way too big.

here is my syslog using cp /var/log/syslog /boot/syslog.txt command

 

Thanks,

Tal

syslog.zip

here is my syslog using cp /var/log/syslog /boot/syslog.txt command

 

Thanks,

Tal

 

I don't think this is the log when you replacing 500GB with 1.5TB because in this syslog there is no entry indicate unRAID was doing disk upgraded.

Whenever doing a disk resizing to a bigger one, you should be able to see following log entries in your syslog. following are from my own log when i upgrade my disk3 to 1.5TB

 

 

Sep  8 17:22:41 Tower kernel: md: disk3 missing

Sep  8 17:22:49 Tower kernel: md: disk3 wrong

Sep  8 17:23:01 Tower kernel: mdcmd (22): start UPGRADE_DISK

Sep  8 17:23:01 Tower kernel: md: recovery thread woken up ...

Sep  8 17:23:01 Tower kernel: md: recovery thread rebuilding disk3 ...

Sep  8 17:23:01 Tower kernel: md: using 1152k window, over a total of 1465138552 blocks.

Sep  9 02:57:39 Tower kernel: md: sync done. time=34470sec rate=42504K/sec

Sep  9 02:57:39 Tower kernel: md: recovery thread sync completion status: 0

 

--------------------------------------------------------------------------------------------------------------------------------------------

your this log however show looks like you don't have clean shutdown and unRAID kick in parity check automatically.

and your parity check is NOT that slow. and there is a warning from unRAID for firmware upgrade, i recall Seagate has

some I/O performance issue, limetech had investigated before, maybe this is the reason?

 

Sep  5 22:44:13 Tower kernel: ata1.00: ATA-8: ST31500341AS, SD17, max UDMA/133

Sep  5 22:44:13 Tower kernel: ata1.00: 2930277168 sectors, multi 0: LBA48 NCQ (not used)

Sep  5 22:44:13 Tower kernel: ata1.00: WARNING: device requires firmware update to be fully functional

 

Sep  5 22:44:13 Tower kernel: ata2.00: ATA-8: ST31500341AS, SD17, max UDMA/133

Sep  5 22:44:13 Tower kernel: ata2.00: 2930277168 sectors, multi 0: LBA48 NCQ (not used)

Sep  5 22:44:13 Tower kernel: ata2.00: WARNING: device requires firmware update to be fully functional

 

Sep  5 22:44:14 Tower kernel: md: recovery thread woken up ...

Sep  5 22:44:14 Tower kernel: md: recovery thread checking parity...

Sep  5 22:44:14 Tower kernel: md: using 1152k window, over a total of 1465138552 blocks.

Sep  6 07:55:07 Tower kernel: md: sync done. time=33051sec rate=44329K/sec

Sep  6 07:55:07 Tower kernel: md: recovery thread sync completion status: 0

 

Is the syslog get cleared everytime you reboot the server? If not, can you send me the command to grab the whole syslog?

Also, I have 3 1.5TB drives and I made sure they were updates before I have installed them into the server around Feb 09.

I guess i can shut down the server and see if there is a newer firmware for my drives and update them one by one.

 

Thanks for all your help.

 

Tal

Is the syslog get cleared everytime you reboot the server? If not, can you send me the command to grab the whole syslog?

Also, I have 3 1.5TB drives and I made sure they were updates before I have installed them into the server around Feb 09.

I guess i can shut down the server and see if there is a newer firmware for my drives and update them one by one.

 

Thanks for all your help.

 

Tal

 

If i recalled native unRAID doesn't save old log (did not keep track of unRAID development so i am not very sure), you need to install extra add-on package. but you can verify it by doing

 

(a) telnet to your server

 

(b) "cd /boot/logs", if you can do this command successfully then proceed to next step, otherwise you don't have historical logs.

 

© "search all logs in this directory for anyone with keyword "UPGRADE".

 

 

 

No luck on the syslog idea. I guess I should figure out how to get the "syslog saver" plugin before I do the next upgrade so I will know what is wrong for the next time I will do it.

I will update this thread when I find the "syslog saver"

 

Tal