Solved!
Go to SolutionModifying Checkpoints time and retention
From within the Celerra Manager GUI there are currently configured 31 daily checkpoints connected to a schedule that runs every day at 03:15 and another 31 daily checkpoints connected to a schedule that runs every day at 03:30.
How would I go about setting up new checkpoints with the following configurations:
6 months of monthlies - 1st of the every month for 6 months a checkpoint is taken and kept for 6 months before starting again.
2 months of weeklies - Every Friday take a checkpoint. After 2 months start again (so write over the first one)
Is it simply a case of using the GUI? Click New on the Checkpoints and create the number of checkpoints needed. Then create a schedule and finally create a mount. Is it that simple? Always thought there might be some command line stuff involved.
I have looked for documentation for clarity, but I have not been able to find any. I know that it is out there, just haven't come across it.
How would I go about setting up new checkpoints with the following configurations:
6 months of monthlies - 1st of the every month for 6 months a checkpoint is taken and kept for 6 months before starting again.
2 months of weeklies - Every Friday take a checkpoint. After 2 months start again (so write over the first one)
Is it simply a case of using the GUI? Click New on the Checkpoints and create the number of checkpoints needed. Then create a schedule and finally create a mount. Is it that simple? Always thought there might be some command line stuff involved.
I have looked for documentation for clarity, but I have not been able to find any. I know that it is out there, just haven't come across it.
847
0
Posted October 7th, 2008 06:00
You are a star. Thank you for coming back so quickly. Now that does sound easier. So, if I wanted 2 months worth of weeklies, would I put 8 (lets just say there are 4 weeks to a month) in the "Keep Checkpoint" box? Obviously I would set it to run every 1 week(s). If I wanted 6 months of monthlies I would put 6 in the "Keep Checkpoint" box.
With the 31 plus the 8 weeklies and 6 monthlies I would only be using 45 ckpt's per file system. Plus, I believe I am right in saying that the checkpoints should be taken at something past the hour. So 07:30 would be ok, but not 07:00.
You can probably tell that I am quite new to this by now. So, is there a quicker way of working out the SAVVOL space than the following -
1. Execute the nas_fs -size command to obtain the SavVol size for each production file system (PFS) that has a checkpoint.
(It is not necessary to calculate additional checkpoints for the same PFS because checkpoints share the same SavVol for each PFS.)
2. Add these SavVol sizes together to obtain the sum of the space consumed by all SavVols.
3. Execute the nas_disk -list command to obtain the sizes of the individual system disks.
4. Add these sizes together to obtain the total space available on the system.
5. Compare the space consumed by all SavVols to the total space available on the system.
The system allows SavVols to be created and extended until the sum of the space consumed by all SavVols on the system exceeds 20% (default) of the total space available on the system.847
0
Posted October 27th, 2008 06:00
Actually I will add more to my last question where I might have the answer. The file system that I have enabled the weekly checkpoint on has a total capacity of 3,249.65 GB. 20% of that is 649.93 GB. So, although the checkpoint is currently only at 127 GB, it can keep on growing to 649.93 GB before I need to make any modifications. Is that right?847
0
Posted October 27th, 2008 06:00
Have had the weekly checkpoints running for 3 Saturdays now. Working excellently. Thank you. A question though. I am monitoring the increase of capacity used just in case. Originally the capacity was 117.2 GB with 31% free when I started. This reduced to 18% free last weekend. On the weekend just gone, the capacity increased to 127GB and now there is 20% free. My question is, how did it automatically increase in capacity? I have looked at the checkpoint properties and noticed the Default Storage Limits. These are all blank, so does this mean that the capacity is basically unlimited until it hits the 20% of the total capacity?168
0
Posted October 27th, 2008 07:00
Excellent. Below info as requested.
Name : emcnas Relocations: /nas
Version : 5.5.28 Vendor: EMC
Release : 1 Build Date: Wed Mar 28 01:54:06 2007
Size : 493647073 License: EMC Copyright
Packager : EMC Corporation
URL : http://www.emc.com
Summary : EMC nfs base install
Description :
EMC nfs base install847
0
Posted October 27th, 2008 07:00
OK, my total storage capacity is 12.1 TB. I have several file systems on it. One of which is the one that I have enabled the weekly checkpoints on. This file system has a capacity of 3.2 TB.
So, 20% of the total capacity is 2.42 TB. Am I on the right track now? It is this 20% that the SAVVOL works with.168
0
Posted October 27th, 2008 08:00
Seriously, you guys rock. Thanks again for your help.847
0
Posted October 7th, 2008 05:00
Its even easier
just create two new checkpoint schedules per file system (one for the monthly and one for the weekly ckpts's)
Any file system can have multiple schedules associated with it.
Just make sure you
- dont try to create ckpts at exactly the same time (at least 5mins between)
- dont exceed the max of 96 ckpt's per file system (less if using Replicator V1)
- have enough SAVVOL space (might want to up the global 20% default and create monitoring events)
The scheduler will automatically create, refresh and mount them
The create new checkpoint and mount piece is only meant for manual checkpoints168
0
Posted October 27th, 2008 07:00
OK, my total storage capacity is 12.1 TB. I have several file systems on it. One of which is the one
that I have enabled the weekly checkpoints on. This file system has a capacity of 3.2 TB.
So, 20% of the total capacity is 2.42 TB. Am I on the right track now? It is this 20% that the SAVVOL works with.
yes, correct - that's the 20% (default) global limit that is set in /nas/site/nas_param
what DART version are you using? (see nas_version -l)168
0
Posted October 27th, 2008 07:00
are you running 5.6.39 or later ?
If yes can you post your fs_ckpt and nas_fs -size output for that file system ?
Rainer_EMC
6 Operator
•
8645 Posts
847
0
Posted October 7th, 2008 08:00
are 4 weeks to a month) in the "Keep Checkpoint" box?
Obviously I would set it to run every 1 week(s). If I wanted 6 months of monthlies I would put 6 in the "Keep Checkpoint" box.
yes, correct. If you want custom names you can specify them also there
yes, thats because by default the NASDB backup runs at the full hour and it locks the NASDB.
If you do need your checkpoints taken at the full hour just ask support to change the time the NASDB backup runs
SAVVOL space than the following -
yes, its much quicker to use the Celerra Manager GUI
Just open the Storage->Volumes Tab and Filter for Volumes of type pool
If you select the lines you want you can right-click "Export Data" and put it into a csv to let Excel do the work
For Total Storage used just go to the Storage->Pools tab, look at the properties of your pools and add up capacities there
Or get create with the -query options (for the non-GUI folks like dynamox
nas_disk -query:servers=1 -fields:name,size -format:"name=%L size=%L\n"
If you keep that many checkpoints around for a long time you most probably need to up that 20% parameter