In this comprehensive guide, we’ll walk you through the process of reclaiming space from Data Domain devices. Learn about how NetWorker manages data retention and how to troubleshoot disk space usage issues with NetWorker and Data Domain For more Information, check our Knowledge Base click here.
Networker reclaiming space from data domain devices, triage guide. In the first section, we'll look to see how networker expires its data networker safe sets reach the end of their retention period. A process called NSR I changes those safe sets to recyclable status. The NSR AM program removes both aborted and expired safe sets once every 24 hours as part of the expiration policy. The expiration works at the same time as the server protection policy in uh inside a network. This is set on a schedule usually to run once a day and it, it is a good idea to only run it once a day if there's no particular reason to run it. More often, you can see here the schedule that it's set to execute on.
There are no different levels for the exploration process. It is just a question of running it and executing it at a time of your choosing as part of this policy. This runs as part of the server protection policy, which is there by default for networker and all new networker installations. Here, we can see the log section. We can have a look at the demon log to confirm that the expiration policy has run, looking for the word expiration in the logs. And we can see here exactly when the expiration policy ran and to see if there are any error messages associated with it, you can see here the exact syntax of what that looks like in the logs when the expiration policy has run. Now, we're going to look at checking networker volumes to be sure that those volumes are in the correct state that they can be treated correctly by networker.
In order to expire data here, we're in the media and dis volume section. This is where we can check a lot of the metadata associated with our volumes to see the current mode of the volume. The name where that volume is located and view the safe sets on that volume. When viewing the safe sets, we can see the clone retention time, which is the time where that safe set instance is expected to expire. So it will be expired after that clone retention time has passed. Provided there are no dependent safes sets involved here. We can also change the expiration or change the status, changing the status from normal or suspect or change the expiration to expire the safe sets now or change the expiration. Should we wish to do so on each individual safe set here, we can see that we can change the recyclable policy from the volume from auto to manual when a volume is marked for manual recycling.
That means that the normal exploration process will not run on that volume. So the default setting is auto. But if we have uh we want to inspect a volume, we have some suspicions about the data on that volume. Sometimes we might want to set this to manual so that the normal exploration process does not run. We need to unmount the volume in order to change the status and then we can change the status over to manual. But this means that the normal expiration process will not run on that volume for while that setting is set to manual. So under normal circumstances, this should always be set to auto so that networker can operate normally on this volume and expire the data. Normally there's a scanned needed flag which can be set here here we're setting it to scan is not needed. Uh scan needed also means that the normal network or exploration policy will not run on that volume when the scan needed flag is set, scan needed is usually used after a disaster recovery process to indicate that networker needs to scan back in the metadata into the networker indexes from these safe sets.
Note also that a volume needs to be mounted in order for the normal expiration processes to run on that volume. Next, we're going to check on the data domain. The actual cleaning of the data from the data domain happens on a schedule defined on the data domain itself. So networker marks the safe sets recyclable or removes the safe sets from the media database. But then it is data domain that will do the actual clearing off of that data from the data domain on a schedule that's defined as the file cy cleans schedule. In the data domain. We can have a look at that schedule by locking into the data domain file cy clean show schedule. And we can see for when that's run and when it has run using files as clean status and when it is set to run on a weekly basis, usually files show space will give us an idea of the clean and data on the data domain. But this is an estimate as there's a lot of d duplication going on on the data domain.
Sometimes it can be hard to calculate exactly how much data is going to be freed up by the expiring of certain safe sets. Next, we'll check the network or logs to see if we can see what has been happening with the expiration policy. We saw previously the expiration policy can be seen through the demon log, the standard demon log of the network or server. We also have an nsrim dot raw log file which is in the standard network or logs directory. We can use the NSR unders render underscore log command against this file in order to put it in human readable format. And for example, we can look for the name of a machine or the name of a safe set to get an indication of what has been happening with the expiration for specific subsets of data. Here, we are looking for the machine VM 3129 and seeing what exploration has been happening for this particular network or client. The uh we can also look for errors or warnings in this log to give us an indication of whether there has been some issues with the run of network or exploration.
There are some advanced parameters which we can use to change the default behavior of network or exploration. The most common of these is the recover space anytime flag which can be set in the NSR debug directory. So you just create a file called recover space any time in NSR debug. And that will have the effect of changing the default behavior of networker to allow eligible safe set files to be deleted even during other read operations on the volume, which is not the default behavior. There are a number of other advanced parameters as well which you can check the documentation for full details. If this under exceptional circumstances, the network or exploration policies functioning needs to be changed. Often when troubleshooting issues with the expiration policy, we will want to run the expiration commands manually in order to see more details around how they're they're operating here using the MM info minus MV command. We can see the list of volumes we have and their different statuses. And then we can run the NSR stage command with the minus capital C minus capital V and volume name. And what this will do is it will find the orphan safe sets and remove them. NSR stage will perform a volume cleaning operation.
It will scan a volume for safe sets with no entries in the media database and recover their space, space for recyclable and aborted safe sets are also recovered from the volume with the safe set entries removed from the media database. This is something that under exceptional circumstances, we may need to run manually, especially during troubleshooting operations. We can also run the NSRIM process itself manually. And there are a number of different switches here. We will be using NSRM with minus capital X which runs the expiration and we can run that against one particular volume. You using the minus capital V and volume name or against one particular client or against one particular client and safe set. So this is something we're going to need to be doing during a troubleshooting process in order to get to the bottom of exactly what's happening with a particular safe set or a particular volume as regards its expiration.
So there are a number of different options we have with this command here, we're seeing NSRM minus C for the client minus capital N for the safe set name and then the minus capital X to say that it wants to run this expiration manually if we're having issues with this and we're getting some problems getting to the root cause of exactly what's causing the issue. We can increase the debug level with the minus capital D, which has different levels here, we're seeing with minus capital D nine. And this will give us a lot more debug data to get to the bottom of exactly what might be causing our issue.
