6,692 thoughts on “CrashPlan packages for Synology NAS”

1. marc April 20, 2012 at 08:19

   Any chance to see a PPC version of the package ? I have the DS410…

   Reply ↓
   1. patters Post authorApril 21, 2012 at 14:54

      Probably not – Java’s the problem. Have a read here:

      Java SE Embedded package for Synology NAS

      Reply ↓
2. catmambo April 21, 2012 at 08:40

   Can’t for the life of me get the Java installation to install correctly. I have the exact same file that the popup tells me =, in the shared ‘public’ folder, yet it keeps saying it can’t see it…what am I doing wrong, or can someone share a link to the correct version which has worked on theirs . I have a DS411

   Thanks

   Reply ↓
   1. catmambo April 21, 2012 at 09:10

      Doh – Yup, I was using Chrome, once I used Firefox it worked a charm. Hope this helps someone else

      Reply ↓
3. bundyo April 22, 2012 at 22:42

   Hmm, seems libffi.so.5 should be linked/copied to /usr/lib after Synology system update or CrashPlan won’t sync (stops watching the selected folders).

   Reply ↓
4. James Berry April 23, 2012 at 08:12

   I have been holding off upgrading to the new DSM (I am on DSM4, but not the new patch which has just come out).

   Does everything just upgrade seamlessly or do I need to re-install crashplan and/or java afterwards?

   Reply ↓
   1. Chris April 25, 2012 at 05:38

      I had to uninstall crashplan and reinstall it. Its syncronising now. Dont know why but it appeared to run for a couple of days but never backed anything up. Just said “Waiting for backup” or something when it was set to always backup. Seems to be working now, will update when its done

      Reply ↓
      1. bundyo April 25, 2012 at 08:12

         The issue is that libffi.so.5 (or a link, didn’t check) is removed from /usr/lib after upgrade and causes the “waiting for backup” message. Symlinking/copying it back should resolve it.
      2. patters Post authorApril 25, 2012 at 11:06

         It’s a symlink – I fixed that in the latest update. The start-stop-status script checks for it at each startup on ARM systems.
5. Bjarne Rasmussen April 24, 2012 at 06:48

   @patters.
   I think this solution is GREAT. If only I could get it working :-)
   I know you have put in a lot of work in this – but you asked me about some details regarding my problem -> https://pcloadletter.co.uk/2012/01/30/crashplan-syno-package/comment-page-3/#comment-3632 … and I hope my answer leads to a new version/script as you mentioned.

   But is there any status on the project – or can I do anything to help / give you more information?

   Kind Regards

   Reply ↓
   1. patters Post authorApril 25, 2012 at 20:47

      Have a try with the new version – it doesn’t assume the existence of /volume1 so hopefully it should work for you now.

      Reply ↓
6. patters Post authorApril 25, 2012 at 20:20

   I have incremented all the builds with quite a few enhancements – see the changelog at the bottom of the blog post.

   The CPPROe client package is updated to version 3.2.1. Don’t upgrade unless your storage provider is running that version on their backend. If you need the older CPPROe packages, they can be downloaded and installed manually from:
   http://dl.dropbox.com/u/1188556/blog/old/crashplanproe20100308-88f8281-003.spk
   http://dl.dropbox.com/u/1188556/blog/old/crashplanproe20100308-x86-003.spk

   Reply ↓
   1. Bjarne Rasmussen April 25, 2012 at 20:53

      GREAT WORK !!
      I had never hoped that a fully functional solution would “pop up” just like that :-)
      This was a really nice surprise – Thank you very much!

      My disfunctional solution – that gave me a message “Wait a few seconds, then stop and restart the package to allow desktop client connections.” simply just needed an update – and then it works,

      Hands down – patters you are the man!

      I owe you a cold one – for sure.

      I hope that everyone else enjoys this as much as me ….. THUMBS UP !!!!

      Reply ↓
      1. patters Post authorApril 25, 2012 at 20:59

         No probs. Cold ones gladly received using the PayPal donate button on the right hand side of this page :)
7. rwojo April 26, 2012 at 00:15

   I’d like to work with you on the x86 inotify issue. Drop me a line at my email on this comment.

   Reply ↓
   1. patters Post authorApril 26, 2012 at 12:27

      I don’t think there’s much that can be done. I’ve just been researching it and I see that inotify support was integrated from glibc 2.4 onwards, but the Intel synos use 2.3.6. The ARM synos have glibc 2.5 which is why they’re fine.

      There is a patch to add it to 2.3.6, but you’d have to recompile glibc which is out of my depth:
      http://www.linuxfromscratch.org/patches/lfs/6.2/glibc-2.3.6-inotify-1.patch

      I tried to compile glibc once and got tied in knots. If someone has the knowledge it could be interesting to try, though I guess it is possible that Synology already built it with this patch included. Perhaps it’s just CrashPlan’s version-matching that is denying us the functionality. It’s odd that the /proc/sys/fs/inotify folder exists on Intel synos if inotify support wasn’t included. Perhaps someone who owns an Intel syno can compile inotify-tools and see if they work.

      Reply ↓
      1. rwojo April 26, 2012 at 16:36

         Before I posted this I did indeed get inotify-tools to compile and run successfully on my Intel DS411+. It works great on glibc =2.4. The options I see are:

         1) Update glibc: I’m not touching it :) I opened a ticket with Synology and I’ll wait. It’s too dangerous to upgrade the entire OS glibc, and it’s just too much work to get it to compile and try with LD_LIBRARY_PATH just for CP.

         2) Spoof the version of glibc from gnu_get_libc_version, add the APIs as necessary using LD_PRELOAD. This could work, maybe I’ll try it.

         3) Hack up CP to remove the glibc checks, but who knows what it would rely on that is missing from glibc 2.3.6 and only exists in glibc 2.4.

         3) Wait for Synology to upgrade DSM 4.x to glibc 2.4+
      2. patters Post authorApril 26, 2012 at 16:47

         I like the look of option 2 also. Keep us posted!
         I also raised a support request, so we may find out the official reason.

         EDIT – just noticed, do some Intel synos have glibc 2.3.6 and others have 2.4 then?
      3. patters Post authorApril 28, 2012 at 14:42

         I got a response from CrashPlan support yesterday, but they didn’t answer my question (which glibc is the minimum requirement?). It was just a “headless systems are not supported” :(
      4. Bjarne Rasmussen April 26, 2012 at 20:35

         @patters
         I’ve got a DS411+II intel syno. How can I check version of glibc so I can add info to your question?
      5. patters Post authorApril 28, 2012 at 14:40

         Run:
         /lib/libc.so.6
         But I’m certain it will be 2.3.6.
      6. rwojo April 26, 2012 at 22:49

         Yay, method #2 works. It will notify me of file changes instantly in the GUI when I modify the filesystem now.

         However, the counts are weird! For example if I just copy and paste the file, the count of Todo files is 1. If I paste again, it goes up to 4? Something must be wrong with the counts in the CP engine/GUI. Can someone verify that it happens on the ARM version on Synology NASes, too?

         I’ve posted the code and binary at https://github.com/wojo/synology-x86-glibc-2.4-shim

         I have to run out for a while, but everything looks great from initial testing. Perhaps you can create a beta package and early adopters can bang on it?
      7. rwojo April 26, 2012 at 22:51

         Interesting, the file count is based on inotify messages for nearly anything it seems, even accessing the directory. Hah weird, probably should notify CrashPlan if it indeed a bug because it artificially inflates the count.

         Of course I’m curious if this happens on the ARM version, too. When I have time I’ll very on my Macs, too.
      8. patters Post authorApril 28, 2012 at 11:47

         I have just found time to check this on mine. I get true counts on ARM. Replacing a file, or accessing it does not increase the count. I thought for a minute that the count did go up to 2 unexpectedly for a single jpg I pasted from a Windows machine, but that was because Windows created a thumbs.db file when I viewed it. Are you sure your testing didn’t have something similar clouding the result (a Mac creating those hidden files in each folder you accessed for instance)?
      9. rwojo April 27, 2012 at 02:33

         Oh, and to your question, no, the comment I posted originally was supposed to say glibc less than 2.4, specifically 2.3.6 works with inotify. I must have just typo’d that because less than or equal doesn’t make sense either :) Does this accept HTML? Tests: < < test
      10. rwojo April 29, 2012 at 03:58

          Counts are fine when I do things with ‘touch’ for example. It must be weird things that OS X does (like the .DS_Store files, etc). I did see a few rapid new files not get counted until the next update, but the count was always eventually spot on.
      11. patters Post authorApril 29, 2012 at 10:32

          Great, that’s good news. On account of how hard all this was getting to maintain, I spent a ridiculous amount of time yesterday rewriting all the scripts and merging them down to just two – using superzebulon’s method, where each script is just a stub that invokes the main one with:
          #!/bin/sh
. `dirname $0`/installer.sh
`basename $0`

          At one point a stray backtick in the main script cost me several hours! The flip side of unification is that a small error in an unrelated area of the script can cause problems in another, and you get no help debugging.

          Now all product versions (CP, CP PRO, CP PROe – on ARM and intel) use the same unified scripts, just with different vars defined at the top. I implemented your suggestions and vastly improved the upgrade process. Logs are migrated, so is the cache (no more rescanning), and you no longer have to log back in with the client afterwards.

          So I just need to incorporate your glibc version shim. I’ll put something up for testing soon.
8. rwojo April 27, 2012 at 03:46

   For NAS units with more than 1GB (I have my DS411+ upgraded to 2GB) it’d be nice to bump the Java heap size to 1GB, something like this in the start-stop-status:

   if [ $RAM -le 128 ]; then
   JAVA_MAX_HEAP=80M
   elif [ $RAM -le 256 ]; then
   JAVA_MAX_HEAP=192M
   elif [ $RAM -le 512 ]; then
   JAVA_MAX_HEAP=384M
   elif [ $RAM -le 1024 ]; then
   JAVA_MAX_HEAP=512M
   elif [ $RAM -gt 1024 ]; then
   JAVA_MAX_HEAP=1024M
   fi

   NAS units usually have a lot of daa on them and this helps CP maintain data structures for such a large amount of files and folders it seems. Perhaps this can be configured somehow in the GUI someday, or able to be overridden from a file that sticks around between upgrades?

   Reply ↓
   1. rwojo April 27, 2012 at 04:38

      Hmm I think I found a bug in the start-stop-status script. My default CP run.conf contains 512m, not 512M. Therefore the sed line isn’t working like it should replacing that value. This should either be made case insensitive or just match the CP default of 512m and replace with 80m, 192m, etc.

      Reply ↓
      1. patters Post authorApril 27, 2012 at 13:58

         I’ll add that 1024 heap to the next version. Are you sure about that 512m value though? The ARM and Intel packages pull the same .tgz installer file from crashplan.com and mine substituted ok to 192M. I can make it case insensitive in future though, just in case.
      2. rwojo April 27, 2012 at 15:34

         Hah, while doing testing I saw it change from ‘m’ to ‘M’. I have no idea why. Probably safe to just match [mM] :)
      3. rwojo April 28, 2012 at 17:57

         After testing, it turns out 1GB adds quite a bit of memory pressure on my box because of what I’m running.

         Do you have a suggestion on being able to configure the max heap amount that would be user configurable and sticky between upgrades? Perhaps a config file that can be placed somewhere?

         For now, I’m moving it back to 768mb or so.
      4. rwojo April 28, 2012 at 18:06

         Also, besides being able to source JAVA_MAX_HEAP from somewhere else, it looks like you can just change the JAVA_MAX_HEAP=xxxMB and rerun the script because sed is expecting to replace 512m to something else. I changed that to the following:

         ed -i “s/-Xmx[0-9]\+[mM]/-Xmx${JAVA_MAX_HEAP}/g” ${SYNOPKG_PKGDEST}/bin/run.conf
9. Einstijn April 27, 2012 at 13:20

   Amazing job and good instructions.

   Instructions work for crashplan cient 3.2.1 (JAVA7) on a DS212+, DSM 4.0-2219. Uploading max 3,0Mbps takes CPU to 90% (setting).

   Is it possible to do any optimizing settings on the DS212+ as well?

   Reply ↓
   1. patters Post authorApril 27, 2012 at 13:41

      The CPU use is caused by the hashing, compressing and de-dupe during backup. Once the seeding is done that will tail off. There’s not much to optimize really.

      Reply ↓
      1. Einstijn April 27, 2012 at 14:20

         I was hoping for better upload speeds with my 50/50 connection then 2-3MBps, that’s all……
10. Guest April 28, 2012 at 08:52

    Thank you for your work. I’m happy with it I’m really am.

    Reply ↓
11. patters Post authorApril 28, 2012 at 17:25

    I believe I have finally solved a longstanding issue for some of you who were consistently reporting that you had to adopt your system every time you restarted the package. Well I got this problem myself after updating to version 007 and it took me a while to figure out.

    CrashPlan will save its .identity file in one of two locations. Firstly it will try to create /var/lib/crashplan/.identity. If that write operation fails, it will fall back to ~/.crashplan/.identity, which is the crashplan user’s home directory (usually /volume1/homes/crashplan).

    It seems that if both of these files exist with different contents, you get this warning everytime the console starts:
    Logged out by authority. Logged in on another computer.

    …and you have to log in and adopt your existing backup records again. My guess is that those of you with /var/lib/crashplan/.identity present are people who had been running a manual install of CrashPlan, since my package never had write access to this directory until version 007. So the fix for this issue will be to delete this folder if it’s present.

    Reply ↓
    1. rwojo April 28, 2012 at 17:49

       I ran into the same issue from manually running the CP engine from the command line (which I don’t do anymore!) while doing testing for the x86 inotify issue.

       Solving it was done by removing one of the .identity files as well, however the two that seem to be tested are /var/services/homes/crashplan/.crashplan/.identity and /var/lib/crashplan/.identity. I don’t see the usage of ~crashplan/.identity in the pecking order in service.log.0.

       Reply ↓
       1. patters Post authorApril 28, 2012 at 17:54

          By ‘~’ I was meaning the daemon user’s home directory (which is /var/services/homes/crashplan).
          Good point about manual startups – I’ll move that logic to the start-stop-status script instead of postinst then, so it deals with that case every startup. Did you see my update about the inotify counts?
       2. rwojo April 28, 2012 at 17:55

          Ah, that makes sense.
       3. Brian Fox April 20, 2013 at 17:50

          This happened to me, somehow the user home service got disabled so it couldn’t store the .identity file anymore.
12. Graham Wheeler April 29, 2012 at 19:20

    Since the latest update, I find that my backups are running really slowly. I have the network bandwidth set to 1Mbps when “user present” and 2Mbps when “user away”, with 50% CPU allowed. I have 25Mbps symmetrical internet connectivity. But I am only seeing about 180kbps transfer rates now, with my backup completion estimated in about 2 years from now. Is anyone else seeing this?

    Reply ↓
13. patters Post authorApril 30, 2012 at 01:13

    New version out now. This one’s got Real-Time Backup support for Intel at last, thanks to rwojo. I also unified the scripts – it was getting too tricky to maintain 6 different versions. The upgrade process is way better too. No more entering your password, or re-scanning. See changelog for more details.

    Reply ↓
    1. Michael Maillot (@Mmaillot) May 1, 2012 at 11:55

       Hi,

       Nice work. Just a quick remark / advice to other users: I used to backup my whole NAS thanks to this Crashplan Package. The backup went almost complete, except for 2 files. Because of those 2 files, the backup would fail every 15′ and try to backup again in 15′. For a reason I do not know – maybe those files are locked in a certain manner – Crashplan cannot backup aquota.user and aquota.group.

       I simply unchecked those 2 files and now everything is fine.

       Michaël.

       Reply ↓
       1. Richard May 8, 2012 at 06:37

          Those 2 files are from the Syno itself. I added them to my exceptions. Otherwise it will try over and over and over and over and over and………
14. Ingmar Verheij May 2, 2012 at 19:46

    Hi,

    Great job! This is way much easier than the installation guides found on the internet!

    I’ve installed CrashPlan Pro (3.2-008, today) and tried connecting from Windows (through an SSL tunnel) but the client (3.8.2010) keeps saying “CrashPlan has been disconnected from the backup engine.”

    The following message is logged in service.log.0

    [05.02.12 21:26:07.576 WARN Sel-UI-R com.code42.nio.net.Connection ] Error building message: Unable to deserialize CommandMessage, IOException in uncompressObject
    [05.02.12 21:26:07.580 WARN Sel-UI-R com.code42.nio.net.Factory ] read() Exception com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=528883553596342675, closed=false, isLocal=true, lat=2012-05-02T21:26:07:574, lrt=2012-05-02T21:26:07:574, lwt=2012-05-02T21:26:03:764, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:51315, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null] , Context@10901009[/127.0.0.1:4243->/127.0.0.1:51315], com.code42.nio.net.Factory$ReadListener@1cc7c50, com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=528883553596342675, closed=false, isLocal=true, lat=2012-05-02T21:26:07:574, lrt=2012-05-02T21:26:07:574, lwt=2012-05-02T21:26:03:764, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:51315, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null]
    com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=528883553596342675, closed=false, isLocal=true, lat=2012-05-02T21:26:07:574, lrt=2012-05-02T21:26:07:574, lwt=2012-05-02T21:26:03:764, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:51315, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null]
    at com.code42.messaging.nio.MessageConnection.buildMessage(MessageConnection.java:283)
    at com.code42.messaging.nio.MessageConnection.enqueueMessages(MessageConnection.java:171)
    at com.code42.messaging.nio.MessageConnection.addMessage(MessageConnection.java:152)
    at com.code42.messaging.nio.MessageConnection.access$000(MessageConnection.java:51)
    at com.code42.messaging.nio.MessageConnection$MessageBuffer.read(MessageConnection.java:754)
    at com.code42.messaging.nio.MessageConnection.read(MessageConnection.java:680)
    at com.code42.nio.net.Factory$ReadListener.processKeys(Factory.java:769)
    at com.code42.nio.SelectorEngine.run(SelectorEngine.java:142)
    at java.lang.Thread.run(Thread.java:722)
    Caused by: com.code42.exception.DebugRuntimeException: Unable to deserialize CommandMessage, IOException in uncompressObject, CommandMessage[null]
    at com.code42.messaging.message.RequestMessage.fromBytes(RequestMessage.java:71)
    at com.code42.messaging.nio.MessageConnection.buildMessage(MessageConnection.java:274)
    … 8 more
    Caused by: com.code42.io.CompressionIOException: IOException in uncompressObject
    at com.code42.io.CompressUtility.uncompressObject(CompressUtility.java:237)
    at com.code42.messaging.message.RequestMessage.fromBytes(RequestMessage.java:66)
    … 9 more
    Caused by: java.util.zip.ZipException: Not in GZIP format
    at java.util.zip.GZIPInputStream.readHeader(GZIPInputStream.java:164)
    at java.util.zip.GZIPInputStream.(GZIPInputStream.java:78)
    at java.util.zip.GZIPInputStream.(GZIPInputStream.java:90)
    at com.code42.io.CompressUtility.uncompressObject(CompressUtility.java:231)
    … 10 more

    [05.02.12 21:26:07.582 INFO Factory$Notifier-UI0 com.backup42.service.ui.UIController ] UISession Ended after less than a minute – 528883553596342675

    Have you got any clues?
    PS: It’s on a DS212+

    Reply ↓
    1. Ingmar Verheij May 19, 2012 at 20:25

       I’ve found the reason why I got the error above. The CrashPlan client installed via your package is 3.2-008 while I’ve got 2010.03.08 installed on my desktop. After upgrading my desktop client to the same version this problem was solved.

       The reason I installed the older version was because this is mandatory for the Dutch storage provider ProBackup (http://crashplan.probackup.nl/). I wrote a blog – http://t.co/2dQLpUg4 – about how to downgrade your package to the version (and configuration) used by them.

       Reply ↓
       1. patters Post authorMay 19, 2012 at 23:17

          A bit further back in the comments I posted the download URLs for the older 2010 version of the PROe package I had made.
15. Robin May 4, 2012 at 13:10

    I cannot see the crashplan package in my list. I can see the Minecraft, Craftbucket, openremote, and the subsonic-backup packages, but no other ones
    I am using a Synolgy DS209+II with DSM 4

    Reply ↓
    1. patters Post authorMay 4, 2012 at 14:23

       Your model has a PowerPC CPU I’m afraid, which doesn’t have a decent workable Java implementation.

       Reply ↓
       1. Sven May 17, 2012 at 00:11

          Hi patters,
          I am receiving the same error (CrashPlan has been disconnected from the backup engine) also on a DS412+ with DSM 4.0 2228.
          Clean install, netstat shows:
          DiskStation> netstat -an | grep ‘:424.’
          tcp 0 0 0.0.0.0:4243 0.0.0.0:* LISTEN
          tcp 0 0 127.0.0.1:4243 127.0.0.1:36612 TIME_WAIT

          I am still using the “old” CrashPlanProe-Version 3.8.2010, because the CrashPlanPro-Server isn`t updated to the new 3.2-version.

          I opened a tunnel (the same way I do with every DS I used before).

          The only difference in my installation is the use of the new Java-Package from Your site and the new version from Oracle:
          ejre-1_6_0_32-fcs-b05-linux-i586-headless-05_apr_2012.tar.gz

          service.log.0 shows the following after starting the CrashPlanClient (GUI) on my Mac:

          [05.17.12 02:07:22.171 WARN Sel-UI-R com.code42.nio.net.Connection ] Error building message: Unable to deserialize CommandMessage, IOException in uncompressObject
          [05.17.12 02:07:22.173 WARN Sel-UI-R com.code42.nio.net.Factory ] read() Exception com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=530941240467324793, closed=false, isLocal=true, lat=2012-05-17T02:07:22:170, lrt=2012-05-17T02:07:22:170, lwt=2012-05-17T02:07:20:588, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:48494, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null] , Context@22823147[/127.0.0.1:4243->/127.0.0.1:48494], com.code42.nio.net.Factory$ReadListener@1fa5e5e, com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=530941240467324793, closed=false, isLocal=true, lat=2012-05-17T02:07:22:170, lrt=2012-05-17T02:07:22:170, lwt=2012-05-17T02:07:20:588, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:48494, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null]
          com.code42.exception.DebugRuntimeException: buildMessage(): Disconnect! Exception! messageId=31689, session=Session[id=530941240467324793, closed=false, isLocal=true, lat=2012-05-17T02:07:22:170, lrt=2012-05-17T02:07:22:170, lwt=2012-05-17T02:07:20:588, #pending=1, enqueued=false, local=127.0.0.1:4243, remote=127.0.0.1:48494, usingProtoHeaders=false, usingEncryptedHeaders=false], dataBuffer=java.nio.HeapByteBuffer[pos=0 lim=10 cap=10], CommandMessage[null]
          at com.code42.messaging.nio.MessageConnection.buildMessage(MessageConnection.java:283)
          at com.code42.messaging.nio.MessageConnection.enqueueMessages(MessageConnection.java:171)
          at com.code42.messaging.nio.MessageConnection.addMessage(MessageConnection.java:152)
          at com.code42.messaging.nio.MessageConnection.access$000(MessageConnection.java:51)
          at com.code42.messaging.nio.MessageConnection$MessageBuffer.read(MessageConnection.java:754)
          at com.code42.messaging.nio.MessageConnection.read(MessageConnection.java:680)
          at com.code42.nio.net.Factory$ReadListener.processKeys(Factory.java:769)
          at com.code42.nio.SelectorEngine.run(SelectorEngine.java:142)
          at java.lang.Thread.run(Thread.java:662)
          Caused by: com.code42.exception.DebugRuntimeException: Unable to deserialize CommandMessage, IOException in uncompressObject, CommandMessage[null]
          at com.code42.messaging.message.RequestMessage.fromBytes(RequestMessage.java:71)
          at com.code42.messaging.nio.MessageConnection.buildMessage(MessageConnection.java:274)
          … 8 more
          Caused by: com.code42.io.CompressionIOException: IOException in uncompressObject
          at com.code42.io.CompressUtility.uncompressObject(CompressUtility.java:237)
          at com.code42.messaging.message.RequestMessage.fromBytes(RequestMessage.java:66)
          … 9 more
          Caused by: java.io.IOException: Not in GZIP format
          at java.util.zip.GZIPInputStream.readHeader(GZIPInputStream.java:141)
          at java.util.zip.GZIPInputStream.(GZIPInputStream.java:56)
          at java.util.zip.GZIPInputStream.(GZIPInputStream.java:65)
          at com.code42.io.CompressUtility.uncompressObject(CompressUtility.java:231)
          … 10 more

          [05.17.12 02:07:22.174 INFO Factory$Notifier-UI0 com.backup42.service.ui.UIController ] UISession Ended after less than a minute – 530941240467324793
          [05.17.12 02:07:22.175 INFO Factory$Notifier-UI0 com.backup42.common.command.CliExecutor ] RUN COMMAND: auto.idle
          [05.17.12 02:07:22.175 INFO Factory$Notifier-UI0 backup42.service.backup.BackupController] UI:: AUTO IDLE… lowBandwidth=0 B/s, activeThrottleRate=20

          Thanks in advance
       2. Sven May 17, 2012 at 01:13

          patters, You mentioned, You don’t have access to an Intel-based Synology-DS.
          If You would like, I can give You access to a brand new DS412+. It is currently not in use. Please contact me via my mail address (merckenssc@me.com), if this would help You.

          I can`t solve the error (losing connection to the backup engine; it doesn´t matter, if I use a tunnel or a modified client (redirected to the IP of the Diskstation).

          Thanks in advance
       3. Sven May 17, 2012 at 01:35

          patters:
          Could it be a problem of the “shown” version?
          the CP-app.log shows:
          CPVERSION = 3.2.1 – 1332824401321
          which is wrong
          I installed 3.8.2010 and this version shows on other installations (older java-package):

          CPVERSION = 1268066820719 (2010-03-08T16:47:00:719+0000)

          Thanks in advance
       4. Sven May 17, 2012 at 01:45

          I think reading helps… especially me ;)
          The new CrashPlanPROe-package-installer downloads the CPClient from the Code42-Webpage. Ok… that isn`t a “wanted” feature or better, their should be an option to install the old version (because the 3.2 (new) and 3.8.2010 (old) aren`t compatible.
          patters: how can I install the old version with You package?
          Thanks in advance
       5. patters Post authorMay 17, 2012 at 07:56

          Look further back in the comments (it’s split into pages). I posted URLs for the old versions.
16. msilano May 4, 2012 at 21:37

    Hi Patters. Excellent work on this – this will inspire me to a) sign up for Crashplan and b) use your affiliate link. Just one small bit of strangeness:

    I installed on a new 412+. Everything is working, but looking at the process listing once Crashplan is up and running, there are…..count em….68 individual processes running, all with the java launch for Crashplan.

    I like multi-threaded programs, just didn’t expect this to be one of them. Each instance is taking 652m of virtual memory according to top.

    I’ve checked all of the launch scripts; they are working properly. Whatever is happening is within the Crashplan engine.

    Is this normal? If not, any suggestions as to how to proceed?

    Thanks,
    Mike

    Reply ↓
    1. patters Post authorMay 4, 2012 at 23:41

       That doesn’t sound right. There should only be one – it’s not multithreaded. Does this carry on when you reboot?

       Reply ↓
       1. msilano May 5, 2012 at 10:59

          Indeed it does.

          I would post the output of the PS command, but it is rather large. There are exactly 64 instances – strange coincidence. Each process looks like this:

          crashpla 649m S N /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPla

          That appears to match the CrashPlanEngine launch for the server. However, there aren’t multiple startups referenced in the engine_output log and the other “instances” aren’t taking CPU time, so I’m not clear as to what is happening.

          Let me know what I can do to help troubleshoot this further.

          -mike
       2. patters Post authorMay 5, 2012 at 12:15

          Hmm strange. Can you try editing /var/packages/CrashPlan/scripts/start-stop-status and remove the LD_PRELOAD bit (the glibc shim) to see if that makes any difference? Had you ever had a manual install of CrashPlan on there?
       3. msilano May 5, 2012 at 12:58

          Can’t seem to reply to the latest comment. Weird.

          In any case, the start/stop script appears to be launching once using the shim. Launching without the shim didn’t seem to make any difference.

          There was an incomplete manual install of crashplan from the synology wiki before I found your site. All remnants of that install were removed. Searches of the file system for any launch scripts or old versions show nothing.

          Thanks again for the reply and the assistance.

          -m
       4. patters Post authorMay 5, 2012 at 14:21

          So if you kill them all then start CrashPlan in Package Center, does it immediately spawn 64 processes or do they take a while to turn up?
       5. msilano May 5, 2012 at 17:45

          Very weird indeed. I can stop and start the processes using CrashPlanEngine directly; in any case, once restarted, they quickly spawn. Not all at once, but after about 1 minute, we have a full allotment of processes. Trying now to run now by just calling the engine once using the same script from CrashPlanEngine; same results.

          -m
       6. brunchto May 5, 2012 at 16:47

          same for me.
          process diseappears as soon as i stop crashplan from the package center. they reappear when i restart it. there’s a root process, the others seem to be child processes

          27847 1 crashpla S N 645m 21.4 0.0 /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan
          27850 27847 crashpla S N 645m 21.4 0.0 /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan
          27853 27850 crashpla S N 645m 21.4 0.0 /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan
          27854 27850 crashpla S N 645m 21.4 0.0 /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan
          27856 27850 crashpla S N 645m 21.4 0.0 /volume1/@appstore/java6/jre/bin/java -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan
          …
17. patters Post authorMay 5, 2012 at 10:33

    Anyone want to help to try and get drive spin down/sleep mode working with this? I’ve had a look at the QNAP forum (which you need to be a member of just to read) and over there someone thought it had something to do with the constant logging. I moved my log folder to /tmp and symlinked it but the drive didn’t spin down as far as I know, though my testing was pretty cursory.

    Reply ↓
    1. eff_cee May 9, 2012 at 03:32

       I thought the lack of sleep was because crashplan continues to open [/volume1/@appstore/CrashPlan/conf/default.service.xml even outside the run between times given ?

       Is this the config file where the the run between times are saved ?

       Maybe CP has to check this file to see if it has been changed ?

       Regardless, it is a pain to have to cron crashplan to get round the issue. Did you get any further on this ?

       Reply ↓
       1. patters Post authorMay 9, 2012 at 08:04

          No, for me it’s kind of a lost cause because Serviio also does this, even with library updates set to manual. Could be a more generic Java problem.
          I guess we could nag CrashPlan support about it, since someone reported that it used to sleep ok. However, they’re pretty consistent in stating that headless operation is not supported.
18. Svein Arild Kristiansen May 5, 2012 at 12:55

    Sorry for the delay, but I have now donated a little £ for you’re great work :)

    Reply ↓
    1. patters Post authorMay 5, 2012 at 13:56

       Thanks very much. :)

       Reply ↓
19. msilano May 5, 2012 at 18:15

    …and the plot thickens….

    The CrashPlanEngine script is indeed launching a single copy of CrashPlan; something within the jar file is triggering the additional launches.

    While this instance was behaving normally, there were another 30 processes spawned. Are we just seeing child processes spawned?

    This is an intel-based ds 412+.

    Thanks again for your help.

    -m

    bash-3.2# ./MSCrashPlanEngine start
    Starting CrashPlan Engine … Using standard startup
    JAVACOMMON /volume1/@appstore/java6/jre/bin/java
    SRV_JAVA_OPTS -Dfile.encoding=UTF-8 -Dapp=CrashPlanService -DappBaseName=CrashPlan -Xms20m -Xmx512M -Djava.net.preferIPv4Stack=true -Dsun.net.inetaddr.ttl=300 -Dnetworkaddress.cache.ttl=300 -Dsun.net.inetaddr.negative.ttl=0 -Dnetworkaddress.cache.negative.ttl=0
    FULL_CP /volume1/@appstore/CrashPlan/lib/com.backup42.desktop.jar:/volume1/@appstore/CrashPlan/lang
    TARGETDIR /volume1/@appstore/CrashPlan
    [05.05.12 14:00:05.974 INFO main root ] Locale changed to English
    [05.05.12 14:00:05.976 INFO main root ] *************************************************************
    [05.05.12 14:00:05.977 INFO main root ] *************************************************************
    [05.05.12 14:00:05.977 INFO main root ] STARTED CrashPlanService
    [05.05.12 14:00:05.980 INFO main root ] CPVERSION = 3.2.1 – 1332824401321 (2012-03-27T05:00:01:321+0000)
    [05.05.12 14:00:05.981 INFO main root ] LOCALE = English
    [05.05.12 14:00:05.983 INFO main root ] ARGS = [ ]
    [05.05.12 14:00:05.983 INFO main root ] *************************************************************
    [05.05.12 14:00:06.222 INFO main root ] Adding shutdown hook.
    [05.05.12 14:00:06.226 INFO main root ] BEGIN Copy Custom, waitForCustom=false
    [05.05.12 14:00:06.227 INFO main root ] NOT waiting for custom skin to appear in custom or .Custom
    [05.05.12 14:00:06.227 INFO main root ] No custom skin to copy from null
    [05.05.12 14:00:06.227 INFO main root ] END Copy Custom
    [05.05.12 14:00:06.239 INFO main root ] BEGIN Loading Configuration
    [05.05.12 14:00:06.365 INFO main root ] Loading from default: /volume1/@appstore/CrashPlan/conf/default.service.xml
    md5 Loaded.
    [05.05.12 14:00:06.601 INFO main root ] Loading from my xml file=conf/my.service.xml
    [05.05.12 14:00:06.713 INFO main root ] Loading ServiceConfig, newInstall=false, version=3, configDateMs=null, installVersion=1332824401321
    [05.05.12 14:00:06.714 INFO main root ] OS = Linux
    [05.05.12 14:00:06.926 INFO main root ] AuthorityLocation@29775659[ location=central.crashplan.com:443, hideAddress=false ]
    [05.05.12 14:00:06.929 INFO main root ] END Loading Configuration
    jtux Loaded.
    ./MSCrashPlanEngine: line 8: 25100 Killed /volume1/@appstore/CrashPlan/bin/nice -n 19 $JAVACOMMON $SRV_JAVA_OPTS -classpath $FULL_CP com.backup42.service.CPService

    Reply ↓
    1. patters Post authorMay 5, 2012 at 23:46

       I have just checked against a syno at my work that’s busy doing its first seed backup (ARM, not Intel), and I can confirm that it has only launched a single process. Since I don’t have an Intel machine I don’t think I’m going to be much use in figuring this out I’m afraid.

       Reply ↓
       1. Richard May 8, 2012 at 06:54

          Whats about the multiple instances on the Intel syno’s?
          (I’m waiting with the update because of this messages)
          What is the impact on the performance or use of the Syno or Crashplan??

          Hope to hear soon!!
       2. msilano May 8, 2012 at 09:13

          I’m leaning towards these being subprocesses as the total memory and CPU usage appears correct for one instance. The file CrashPlanEngine.PID (as per the launch script) contains 1 entry. And Crashplan listens as expected.

          Just my $.02.
       3. patters Post authorMay 8, 2012 at 10:23

          I don’t have an Intel syno so I can’t really help. However, there is no fundamental change to how CrashPlan is started in my newer package version (apart from the Intel glibc shim – which msilano has confirmed doesn’t cause this behaviour). So, is there a chance that this has been happening all along?
       4. Richard May 8, 2012 at 10:49

          I did the backup 15 min’s ago.
          No problem what so ever ;-)

          No error messages etc…

          But I don’t have the feeling the realtime scan is working… although I don’t get an error message…..
       5. patters Post authorMay 8, 2012 at 11:26

          You can check whether real-time backup is working by opening the console and then changing or adding a file. You should see the console report it, and schedule it for the next backup in 15mins.
       6. Richard May 8, 2012 at 12:58

          I tried that indeed, but it does not report it…
          Will try it again when getting back home.
       7. patters Post authorMay 8, 2012 at 13:00

          rwojo tested it on his DS411+. Which model do you have?
       8. Richard May 8, 2012 at 13:05

          I’m using: 1812+
       9. Richard May 9, 2012 at 06:09

          Tried again..

          -Waited till the backup run finished
          -Added 10 documents
          -waited 45 minutes
          -no automatic selection scan :(

          I’ll stick to my hourly schedule… it works

          backup freq: 15min
          verify selection every: 1 hours
       10. patters Post authorMay 9, 2012 at 08:06

           Any errors in the engine logs in /volume1/@appstore/CrashPlan/log?
       11. Richard May 9, 2012 at 08:17

           5 similar errors in thread 21, 25, 26, 32 and 33

           Exception in thread “Thread-21” java.lang.NullPointerException
           at com.code42.jna.LinuxPlatform.isSymlink(LinuxPlatform.java:271)
           at com.code42.jna.inotify.JNAInotifyWorker.depthFirstTraversal(JNAInotifyWorker.java:89)
           at com.code42.jna.inotify.JNAInotifyWorker.depthFirstTraversal(JNAInotifyWorker.java:116)
           at com.code42.jna.inotify.JNAInotifyWorker.depthFirstTraversal(JNAInotifyWorker.java:116)
           at com.code42.jna.inotify.JNAInotifyWorker.run(JNAInotifyWorker.java:52)
           at java.lang.Thread.run(Thread.java:662)
       12. patters Post authorMay 9, 2012 at 09:18

           Hmm. The same error was observed on ARM platforms when the JNA library wasn’t present:
           http://forum.qnap.com/viewtopic.php?p=235396 (you may have to register in order to see this).

           @rwojo – any ideas? I wonder whether a newer JNA lib might fix it.
20. Jim May 9, 2012 at 15:04

    I have a DS212+ that i installed this on, and i did the recent update to it the other day, and it does not seem to hold the network settings so if i change it to us my max upload speed it does not go it seems to hang out at the 300 setting.

    Reply ↓
    1. Jim May 9, 2012 at 15:06

       my Max upload should run around 100 KB/s and it is hanging around 30 KB/s

       Reply ↓
    2. Jim May 11, 2012 at 15:30

       I discovered that having it generate thumbnails and do a back up at the same time does not work so well, once i stooped doing the thumbnails the speeds for crash plan leveled out to what i expect them to be.

       Reply ↓
21. david May 9, 2012 at 23:03

    Is there any way to provide better instructions? I just have a new DS212j that I buy only for CrashPlan and I am stuck on the java part. The package installer says that I need to manually download it and place it to the “public” folder but I issue these commands:

    cd /
    find * | grep public

    Only to discover there is no public folder. I enable SMB and figure out how to login again to discover there is no public folder. I would love to create the public directory… but where? /public? /volume1/@tmp/public? It is very hard to follow without a detailed instruction and I hope you understand.

    So if please you can provide better detailed instruction like “place it in /var/tmp/public” that would be great!

    Reply ↓
    1. patters Post authorMay 9, 2012 at 23:25

       Make a new top level folder in DSM’s File Browser and call it ‘public’. That will be shared automatically, so put the downloaded file in there.

       Reply ↓
    2. Jim May 10, 2012 at 00:35

       you need to make sure that the Webdav service user under system internal users has access permissions to the public folder as well

       Reply ↓
       1. patters Post authorMay 10, 2012 at 08:38

          Really? Wasn’t a problem for me on a four bay syno at work, though I did add full access permissions for everyone – which is how this folder is defaulted on single bay NAS systems.
       2. Jim May 11, 2012 at 15:33

          yes that is what i had to do to get it to work correctly
22. David May 11, 2012 at 01:28

    Does anyone know if this works with the DS212+ (513MB DDR3), or is 1GB of NAS RAM a firm requirement?

    Reply ↓
    1. Jim May 11, 2012 at 15:32

       I am using on a DS212+ and it works great! however make sure that when you are doing your initial backup or large backups to crash plan you are not taxing the CPU with making thumbnails for photo station

       Reply ↓
       1. patters Post authorMay 11, 2012 at 16:06

          Code42’s official requirement for CrashPlan is 512MB, but even that allows for very large backup sets. My package sizes the Java heap size appropriately for the available RAM in an effort to make sure it didn’t attempt any kind of suicidal paging to disk while backing up. Given that most home users are constrained by bandwidth I’m guessing it’s not practical for most people to backup the whole NAS. I’m only protecting around 60GB of mine, for which a Java heap of 192MB is perfectly adequate. However, at work I’m uploading 3TB of data using CrashPlan PROe from a RS411 with the same heap size. In fairness it is made up of mostly very big files though which probably reduces the overhead somewhat (master edits of videos).
       2. David May 14, 2012 at 21:29

          Thanks so much — seems like a great home NAS/backup solution; the best combo I’ve seen. I’ll probably be back on this thread soon while setting mine up!
       3. Chris June 14, 2012 at 17:18

          I don’t want to backup things off my DS212+ but I do want to backup to it. I would want to have several computers backup to it. Maybe 1-1.5 tb total. Anyone have experience with this? Any performance feedback would be great.
23. Ulrik May 14, 2012 at 07:19

    Just upgraded from DS508 to DS1812 last night, and now want to add Crashplan from you repo. But it fails out during installation with “There was a problem downloading CrashPlan_3.2.1_Linux.tgz from the official download link, which was….(the link is correct)”. What could be wrong?

    Reply ↓
    1. patters Post authorMay 14, 2012 at 09:49

       It just uses wget so maybe there was a temporary glitch with the CrashPlan.com website. I just checked and it works for me now. Is it still failing?

       Reply ↓
       1. Ulrik May 14, 2012 at 09:55

          Yes, it is still failing. Is there a log somewhere to give me an idea of what goes wrong? I can download the file manually – also with wget from a SSH connection to the DS1812.
       2. Ulrik May 15, 2012 at 17:43

          OK, I got it installed now. The install failed earlier because of a disk expansion. But now I got another problem. I have stopped and restarted the CrashPlan addon on ny Syno, and edited the ServiceHost line in ui.properties, but I cannot connect with my new CrashPlan account.
24. Rogier May 14, 2012 at 21:00

    I’ve installed the Java 7 package on my DS411j in order to be able to use Crashplan. Everything is working fine, however, after some hours, Java is consuming all CPU power. Result is that I can’t connect anymore to the Crashplan installation and the NAS becomes very slow. Currently, the only remedy is to restart the NAS once a day. Does anyone else seems to have the same issues? Does anyone how to solve this? Any help is appreciated!

    Reply ↓
    1. MJ June 13, 2012 at 20:40

       I had the same problem on mine. I’m thinking it just doesn’t have enough RAM to handle the client. I plan to upgrade to a DS411 or 411+

       Reply ↓
25. Matthew May 15, 2012 at 20:31

    Hi we are currently trying to set up crashplan pro e on a DS1512+ and we cant seem to get the client installer to find our server on the synology nas box.

    If anyone has gotten this working and could lend some assistance we would gladly pay a consulting fee.

    Mathew@northbaytek.com

    Reply ↓
    1. patters Post authorMay 15, 2012 at 21:07

       The naming is a little confusing. My CrashPlan PROe Synology package is only a PROe *client* (as titled Package Center) – i.e. you’d still need to pay a provider for storage hosting then use a client on a PC to get the syno to connect to that provider. That’s how I’m currently using it, connecting to http://crashplanuk.com. You don’t need to hack any files, you simply enter the connection URL into the GUI and log on when the client first connects to your syno.

       PROe server is what the storage provider would run on their hardware.

       Reply ↓
       1. DJ Forman May 15, 2012 at 21:13

          I thought you could have the ProE server running on the Synology box and use that as the storage provider?
       2. patters Post authorMay 15, 2012 at 21:22

          You might be able to, but you would have to buy master keys from CrashPlan which I would expect to be quite expensive given that it’s basically aimed at the datacenter market.
       3. DJ Forman May 15, 2012 at 21:30

          Thanks patters, we bought 40 perpetual licenses with 1 year of support. But of course they won’t support this implementation, nor will Synology. Not directly anyway.

          I friend of mine has this same setup deployed on his Synology but he had to get help from an outside source to get it completed. I’m hoping to get him to send me the instructions he follows.

          In fact his implementation is more complicated than mine because he uses 2 Synology boxes, one on each seaboard, that replicate to each other.

          Fingers crossed. If you are into a quick consulting gig to help us get this setup that would be great. Otherwise will stumble along trying to make it work.
       4. patters Post authorMay 15, 2012 at 21:40

          I see. I could be interested, however it’s already 22:38 here in London and I’m only just leaving work! Backup Exec 2012 migration and iSCSI issues…
          If by some miracle I’m feeling up to it in the next few hours I’ll let you know.

          The only problem about outside assistance for CrashPlan implementation is that it will almost certainly fail and need additional work once a new version is released – which is why I made the packages. I guess if it’s not complicated I could end up making a PROe server package. Thing is, I can’t experiment without a master key.
       5. DJ Forman May 15, 2012 at 21:47

          Backup Exec issues, bleh. Sorry for your pain. I moved to Acronis and/or StorageCraft and never looked back. I’m open to tomorrow morning your time also. I’ll stay up late on my side. I’ve got several clients in the UK and Germany that I need to work with tonight so I’ll already be up.

          I’m going to try to install the ProE server manually as if the Synology were a “real” Linux server. Supposedly this should work. By the way, I’m not cheap, I’m sure I can make some consulting worth your while.
26. DJ Forman May 15, 2012 at 21:10

    We got the Crashplan ProE headless client installed on a Synology NAS. It looks like it’s listening on 4243. But no matter what I try I can’t get the Crashplan Windows client to connect. It just spits out an error about not being able to connect to port 443 on the IP of the NAS.

    I changed ui.properties, but not sure if I edited it properly. It’s just a single line with fields seperated by the # sign. I entered to additional lines to the bottom to point to the server and port. No luck.

    Could use some help. Happy to pay for some consulting. I need this up and running tonight so I can deploy 40 users. I’m pretty sure it’s just a minor mis-config.

    djforman1(at)yahoo.com

    Reply ↓
    1. patters Post authorMay 15, 2012 at 21:28

       Hmm this comment above isn’t threaded properly so it’s going to be confusing to read…

       As per the posts above, PROe client can’t really be used like this. I would guess that to achieve what you’re aiming to do, you might be able to use the normal CrashPlan package on the Syno then run the normal CrashPlan clients on your 40 computers. Then look at the ‘backup to a friend’ option, taking the friend code from the syno’s CrashPlan instance.

       I would guess that there is an upper limit on the number of friend connections though, and 40 is likely to be on the high side.

       Reply ↓
27. OjaSapNL May 16, 2012 at 19:13

    Hello,

    For everybody that hates that your Synology harddrives don’t hibernate anymore I have created an solution. First I was using the Crontab solution, but that didnt fit. If i have an large backup and it dont fit the time window the backup will not be completed, and if u have an large one your harddrives stay online to long.

    I created an Python script that reads the CrashPlan logfile every 5 minutes. In the script you fill in the time that the CrashPlan service need to be started. Then it starts it, and reads every 5 minutes if its completed. Then It sends an e-mail with the lines of the logfile, and stops the CrashPlan service. I’am currently testing it. If anybody is interested, tell me and I will give u the script.

    OjaSapNL

    Reply ↓
    1. TopL June 4, 2012 at 17:17

       Hey OjaSapNL,

       I’m interested. I’m whittling down why my NAS isn’t hibernating and isolated it down to CrashPlan. If you can provide details on how your script works, that’ll be great!

       You can ping me at (please replace the splats accordingly): tl.2012 * xemaps * com

       Thanks!
       TopL

       Reply ↓
28. Ronald Rademaker May 17, 2012 at 06:32

    Hi, I’m intrested. Can you pass me the script ?

    Thx,
    Ronald

    Reply ↓
    1. OjaSapNL May 17, 2012 at 10:31

       I’ve posted it on my blog. Please let me know if it works for you.

       http://www.ojasap.nl/2012/05/17/crashplan-scheduler-script/

       Reply ↓
29. Ronald Rademaker May 18, 2012 at 07:57

    Hi,

    How can I check the status of my Crashplan in a command shell ? Which location/command is needed to be used?

    Confused as I also Have my old install which give me the error:

    DiskStation> ./crashplan status
    Could not find JAR file /opt/crashplan/bin/../lib/com.backup42.desktop.jar

    The new one using this package can be stopped and started in the Package Center without issue and backup is working properly

    Thx,
    Ronald

    Reply ↓
    1. patters Post authorMay 18, 2012 at 08:18

       From memory, I think the CrashPlan launcher script (which my package’s start-stop-status script invokes) expects you to be in the program folder since it contains relative paths. I don’t think their launcher has a status function so I would suggest you try this:
       cd /volume1/@appstore/CrashPlan
/var/packages/CrashPlan/scripts/start-stop-status status && echo running || echo stopped

       As you can see, packages themselves are in the @appstore folder, but the metadata from Package Center and the package scripts end up in /var/packages.

       Reply ↓
30. Darcy May 23, 2012 at 16:55

    Fantastic work creating this package! I’ve been running some test backups on a 30-day trial, and it’s been great. Will sign up shortly with your affiliate link :)

    Reply ↓
31. Andrew Stuckey May 24, 2012 at 13:50

    Hey Patters. Thanks so much for your work on this. Hope you can help us out with our setup.
    Will definitely give a donation if you can help us get our backups working ;). There you have that in writing!

    We bought a DS411j as our main data storage and file server. We also bought a Drobo (DAS version) to use as remote backup of the NAS using Crashplan. The plan is to seed the initial backup with the Drobo connected directly to the NAS, then move the Drobo to a friend’s house and continue with incremental backups over the internet.

    So far we’ve managed to install your CP package on the DS411j and have successfully run some test backups to the Drobo while connected directly to the NAS via USB. To do this we’re running the CrashPlan client on an iMac on the same LAN as the NAS by changing the host IP.

    Then in order to test the seeded backup over the network we unplug the Drobo from the NAS and connect it to the iMac. However CrashPlan won’t allow us to select a network drive as the backup destination. In fact, no networked machines or drives even appear in the list. We’re only able to select a local folder on the NAS or the Drobo when connected directly to the NAS.

    Are we going about this the wrong way? Have we missed a step?
    Very disheartening to have come this far only to be stumbling at the last hurdle.

    Any help would be very appreciated.
    cheers

    Reply ↓
    1. patters Post authorMay 24, 2012 at 15:23

       You would need to get the iMac CrashPlan GUI client connecting to the Backup Engine instance running on the iMac (so undo your conf/ui.properties modification). This iMac would then be acting as a separate CrashPlan setup (like the one you want to run at your friend’s house). I haven’t tried this, but I would imagine that you would get the friend code from that setup. Then I think you would plug in (on the iMac) the Drobo you already seeded from the NAS, and in the advanced backup destinations options (I’m not near a CrashPlan GUI to check) you can set the actual folder on disk where the friends backups are saved. I think you just attach the seeded folder that’s on the Drobo (it should be a local drive since it’s DAS and it’s plugged into the iMac). Then I’m guessing that you would once again edit ui.properties to switch the GUI client back to connect to the Synology and add a backup destination using the friend code. Hopefully it should just be aware of the existing seed backup, since the GUID will be recognised. This is all untested though. I personally just pay for CrashPlan+

       Reply ↓
       1. Andrew Stuckey May 25, 2012 at 01:05

          Thanks. We considered CrashPlan+ but with > 4TB of data to backup and being in Australia we unfortunately can’t use CrashPlan’s seeding service, which makes it very impractical. To do the initial backup over the net would probably take several months!

          I’ll play around with the CP client on the iMac and see if we can get it to work… stay tuned.
       2. Andrew Stuckey May 26, 2012 at 01:27

          Thanks Patters. We seem to have got it working!
          I seeded a small backup, then moved the external drive to a second Mac (on our CP account) and added the backup archive to the Mac through the CP client. It instantly recognised that the backup had come from the NAS and prompted to start synchronising. Great. Lets hope it works over the internet to a friends computer once we’ve finished the initial seed.

          Resuming the seed backup…

          New issue…

          It’s backing up at ridiculously SLOW speeds and I’m having trouble isolating the bottleneck.

          Our setup is the DS411j running 4 x new 3TB Seagate Barracuda 7200rpm drives connected to a Gigabit Apple Airport Extreme. I’ve stopped all packages except CrashPlan and your Java plugin. There are also no Time Machine backups or any other processes running (as far as I’m aware).

          With the backup drive directly connected to the NAS via usb2 it’s transferring data at an agonising slow 6-9 Mbps (a far cry from the theoretical 480 Mbps of usb).

          With the backup drive connected to the Mac (which is then connected to our Gigabit Airport Express by Cat6 cable) we get even slower speeds of around 3-5 Mbps or an average of roughly 600 KB/s looking at the DSM resource monitor.

          So not even 1MB/s which is completely useless and a very far cry from the 29-85 MB/s benchmarks posted on the Synology site http://www.synology.com/products/performance.php?lang=enu.

          I haven’t done any other performance tests at this stage other than via CrashPlan, so I don’t have anything to benchmark it by. But do you have any immediate suggestions? With no other processes running could it be an issue with CrashPlan?

          Could it possibly be an AFP or SMB issue? Although this would not explain the terrible speed when connected directly to the NAS usb port.

          There also appears to be a common speed issue between Synology and Macs as documented here (http://forum.synology.com/enu/viewtopic.php?f=14&t=34172&hilit=SLOW). Maybe this might shed some light on the issue.

          Any help specifically with identifying potential problems running CP on the NAS would be much appreciated.

          cheers
          Andrew
       3. patters Post authorMay 26, 2012 at 10:01

          Try disabling de-dupe and compression in the advanced backup options. For the ARM CPUs I think it’s a big performance hit. I hate to say it, but I think your DS411J won’t have enough RAM for such a large backup set. The default CrashPlan heap is 512MB and for backing up several terabytes, Code42 support have advised several people to use 1024MB. Only the Intel Synology models have this much RAM (Expandable to 3GB I think).
       4. Andrew Stuckey May 26, 2012 at 02:49

          Have now tested the external drive using FW800 and unfortunately speeds are still the same dreary 3-5 Mbps, so we can probably rule out the usb as the bottleneck.

          For all test so far NAS CPU has only averaged 40% and RAM 60%.
       5. Andrew Stuckey May 26, 2012 at 13:13

          Where are the advanced backup options? In the DSM or CrashPlan? Can’t find them.

          The RAM issue you’ve mentioned with the ARM machines seems stupid. Does this only apply to CrashPlan? And if so why does CP suck so much RAM? or are most other backup applications the same?

          There must be a viable remote backup solution for the Non-Intel models. If not Crashplan then what else?
       6. patters Post authorMay 29, 2012 at 22:16

          CrashPlan GUI -> Settings tab -> Backup -> Advanced settings.

          CrashPlan uses RAM for keeping track of files’ checksums, block hashes, and other metadata I would guess. With several terabytes in your backup set, it’s not difficult to see how even a pretty efficient engine would need a fair amount of RAM. There is a cloud option among the official Synology apps (HiDrive), though I have seen people commenting on its lack of reliability. I have amended the Notes section to draw attention to the RAM issue in case people are shopping for Synos with the express intention of using CrashPlan for large backup sets.

          I agree it is a shame that the ARM models don’t have a DIMM slot, but then they are very cheap embedded systems that are primarily designed to do one thing, serve files – not run applications. That’s a bonus.
       7. Andrew Stuckey May 30, 2012 at 02:14

          Meh… the Advanced settings are disabled on the Free plans. Seems like a CP+ feature. oh well. But you’re right about the RAM issue, as soon as I launch CP my whole system chokes. “they are primarily designed to do one thing, serve files… not run applications.” Wish I knew this before I ordered!

          So what I’m thinking now is to buy an Syno Intel NAS with more RAM, sell the brand new Drobo and keep the DS411j as the offsite backup machine. Which model Syno would you recommend buying? It’s not clear from the specs on the Synology site which models are Intel vs ARM. We’ll need something with enough grunt to run 3-4 applications simultaneously including TimeMachine and offsite backups without compromising the NAS performance. Clearly the DS411j is not up to the job which is disappointing. Will it be fine as a backup unit?
       8. Andrew Stuckey May 30, 2012 at 02:17

          Another thought…when did Synology start using Intel processors? If we don’t want to spend a heap, would be cost effective to buy a 2010-11 model with the power we need? Happy to consider older models.
       9. Andrew Stuckey May 30, 2012 at 04:39

          Just found this :) http://forum.synology.com/wiki/index.php/What_kind_of_CPU_does_my_NAS_have
          So it would seem that we are best off looking at either a DS710+, DS411+, DS1511+ or the DS412+.

          What would you recommend?
       10. Andrew Stuckey June 4, 2012 at 12:51

           Thanks for your help Patters.
           Donation made :)
       11. patters Post authorJune 7, 2012 at 07:44

           Thank you!
32. Samuel Manso (@samukas_m) May 24, 2012 at 13:52

    Hello patters! Recently I had a problem with my crashplan installation not backing up and I was told by crashplan to:
    “2. Edit the below line in /usr/local/crashplan/bin/run.conf
    3. Find this line (near SRV_JAVA_OPTS): -Xmx512m
    4. Edit to something larger such as 640, 768, 896, or 1024. E.g.: -Xmx1024m”

    I did that, and I also changed the value set in “/volume1/@appstore/CrashPlan/syno_package.vars” as specified in your post.

    It seems to have worked, but I’m wondering if it was really needed to change the value in both, or if the the “vars” file would be enough.

    Anyway… a little donation coming your way (wish I could give more)… but it’s well deserved! Thanks to your package I’m backing up 3TB online to crashplan.

    Reply ↓
    1. patters Post authorMay 24, 2012 at 15:11

       I’m assuming you do actually have more than 1024MB of RAM in your syno? If not then the performance may get pretty bad as it does loads of paging to disk. Are you running on Intel?
       Oh, and you only need to change the value in syno_package.vars – that value will override the one in run.conf. This will setting survive a package version upgrade.

       Reply ↓
       1. Samuel Manso (@samukas_m) May 24, 2012 at 16:03

          I switched the value to 768 at the moment just to try it out (also turned off all my other packages besides crashplan).
          I’m running an Intel (DS1511+) with only 1GB but I already bought an extra 2GBs, so I’m thinking I’ll give either 1 or 1.5GB to crashplan and the rest to the system. What do you think? Also, for example, to give 1.5GB, should I write 1536 on the the file? I’m thinking that’s the right number.

          Glad to know that I only need to change the value in one file, and that it will survive upgrades :)
       2. Samuel Manso (@samukas_m) May 24, 2012 at 22:39

          Definitely worth the upgrade! :D
          I’m now running with 3GB of RAM, changed the value to 1536 on syno_package.vars and at this exact time, “java” is using 900MB of resources. Of course 1GB was not enough :)
       3. Charlie June 6, 2012 at 02:08

          Can someone please advise on how to change the values in the .vars file? When I do a vi to it, it appears to be a blank file. I am no command-line whiz, but can “eventually” get around. Just don’t know how to add/ edit this line because I am restarting the service about 3x per day (initial 2.5tb backup on DS1010+ w/ 3GB ram)
          Thanks,
          Charlie-
       4. patters Post authorJune 7, 2012 at 07:40

          It shouldn’t be empty. Are you logged in as ‘root’? It has the same password as your admin account. That file should look something like this:
          ~$ cat /volume1/@appstore/CrashPlan/syno_package.vars
#uncomment to expand Java max heap size beyond prescribed value (will survive upgrades)
#you probably only want more than the recommended 512M if you're backing up extremely large volumes of files
#USR_MAX_HEAP=512M

          LIBFFI_SYMLINK=YES
MANIFEST_PATH_SET=True
~$
       5. Charlie June 7, 2012 at 12:04

          Hi Patters,
          I tried ssh and telnet as root and when I do a vi to that file it is empty. I quit without saving and tried to do a cat as you show in your reply and I get a “no file exists…”.
          I know it is running and the JRE6 is installed (although it says stopped with no way to start it). Is something else improperly installed perhaps or could it be in another path?
          If this helps, I was running CP+ then I uninstalled that service and installed the CPpro version. I upgraded my CP account to try and get more throughput. It seems to be better (when it runs).
          Thanks,
          Charlie-
       6. patters Post authorJune 7, 2012 at 17:09

          Have you tried creating that file by running:
          echo USR_MAX_HEAP=768M > /volume1/@appstore/CrashPlan/syno_package.vars

          If that doesn’t work, perhaps your NAS isn’t using /volume1 for its appstore. Try:
          ls /
       7. Charlie June 7, 2012 at 17:16

          Patters,
          Well, that yields a directory doesn’t exist:

          DiskStation> echo USR_MAX_HEAP=768M > /volume1/@appstore/CrashPlan/syno_package.
          vars
          -ash: can’t create /volume1/@appstore/CrashPlan/syno_package.vars: nonexistent directory
          DiskStation> ls /
          bin initrd lost+found sbin var
          dev lib mnt sys var.defaults
          etc lib64 proc tmp volume1
          etc.defaults linuxrc root usr
          DiskStation> cd ..
          DiskStation> ls
          bin initrd lost+found sbin var
          dev lib mnt sys var.defaults
          etc lib64 proc tmp volume1
          etc.defaults linuxrc root usr
          DiskStation> cd /volume1
          DiskStation> ls
          @afpd.core @postfix DataFiles aquota.user music
          @appstore @spool Documents crashplan photo
          @autoupdate @tmp Photos downloads public
          @database ATV1 Time Machine homes video
          @eaDir ATV2 aquota.group iTunes
          DiskStation>

          But, I cannot cd to @appstore – it says not found, so I don’t know if the CrashPlan directory is beneath it.

          Thanks,
          Charlie-
       8. patters Post authorJune 7, 2012 at 19:17

          and what’s in @appstore?
       9. cfpsystems June 7, 2012 at 19:18

          FIGURED IT OUT!!
          Had a buddy look at it and (disclaimer that I am not a command line wiz), found that I was trying to change directory with a “/” in front on @appstore.
          Once in, found that the file I needed to update was under the directory CrashplanPro. Heap size adjusted and enabled – will report back after its run a while.
          Thanks,
          Charlie-
       10. patters Post authorJune 7, 2012 at 20:35

           Glad you got it figured out. I forgot that you had mentioned it was CP Pro, and I have been quite slow to get back to you – been very busy in real life.
33. devalnornor May 29, 2012 at 20:56

    Your Crashplan Package do not appear in Package Center. Java 6, 7 & other stuff are there but no Crashplan :'(

    Reply ↓
34. Adam May 29, 2012 at 21:24

    Great package, thanks for putting it together! I’ve run across an issue where after uninstalling the package from my DS1511 using package center, the package is no longer listed in the repository for installation. Any ideas how to make it reappear?! (This is not the window size bug in DSM 3.2, I only see three packages to install)

    Reply ↓
    1. patters Post authorMay 29, 2012 at 21:35

       Sorry, I had just updated the package and my NAS went a bit weird earlier on (no Java apps would start properly). I took the new packages down just in case. However removing Java, rebooting and re-installing Java fixed it so they’re back on the repo now.

       Reply ↓
35. Scott May 30, 2012 at 06:36

    Thanks for the package, I was looking at going through your link to try out CrashPlan Pro w/t my 1812+ but when i followed your instructions to add your package to the package center but the CP Server is showing version 3.2? Is this upgradeable to the new v4.0?

    Reply ↓
    1. patters Post authorMay 30, 2012 at 07:16

       Where do you see CrashPlan 4.0?

       Reply ↓
       1. Scott May 30, 2012 at 07:42

          Sorry, maybe I was mixing up a PROe e-mail I got a work, 3.2 looks like the current server version for the ‘business’ hosted version I was looking to install on my synology at home.

36. Pingback: Ingmar Verheij – The dutch IT guy » Backup Synology to CrashPlan Pro (on Dutch server at Pro Backup) » Ingmar Verheij - The dutch IT guy

37. Jemima June 14, 2012 at 10:21

    Hi Patters – This is great, and I hope I can get it to work. I’ve installed Java 7 per your instructions, but the Package still gives me a ‘Java is not installed or correctly configured’ error. I’ve tried rebooting the NAS, but still no joy. I’m running as DS209. Can you offer any pointers?

    Reply ↓
    1. Jemima June 14, 2012 at 10:42

       Some more info… I am not convinced Java is running correctly – I followed the instructions and the package install completed, but I have the following in the log window of the package:

       /var/packages/java7/scripts/postinst: line 28: java: not found

       Systems installed locales:
       C
       en_US.utf8
       POSIX

       JAVA_HOME=/volume1/@appstore/java7/jre
       TZ=Europe/Brussels

       Reply ↓
    2. Jemima June 14, 2012 at 11:02

       Yup, forgive me. Java 7 not working for some reason, but I’ve now installed Java 6 without issue and package now installs :-)

       Reply ↓
38. John K June 14, 2012 at 16:21

    So my DS411J clearly isn’t up to the task of running CrashPlan. I’m curious what models other people have had success with ( I will probably need to eventually handle at least 2TB+ of data). I am looking at the DS411+II which comes with 1 gig of RAM, it’s $200 more to step up to a model that has an extra RAM slot and more than I really want to put into this, just for my personal needs.

    Reply ↓
    1. Joe June 15, 2012 at 13:22

       Hi John,

       I’m happily running Crashplan on a 6TB DS411J. I’ve had to tweak the memory usage (run.conf) to 1024M. As you know, the DS411J has only 128M of physical memory so there’s a performance hit with paging.

       Reply ↓

39. Pingback: Anonymous

40. Pingback: Synology DiskStation und CrashPlan | Alexander Benker

41. Pingback: Confluence: Knowledge Base

42. Ulrik June 21, 2012 at 07:47

    Problem: CrashPlan package indicated an update was available, and I let it upgrade. It failed when downloading the installer from Code42 website, and the result was, that the package was now removed! So I downloaded it manually and placed the installer in the \public share and then installed the package again. This time it installed with success, BUT when I then connected the CrashPlan client, the Synology was listed as a new CrashPlan engine (with green indicator) and the old engine (with same name) had a grey indicator button. Obviously my CrashPlan license was also gone. It now says I will loose all of my backups, if I transfer the license to the new CrashPlan engine… what to do???

    Reply ↓
    1. patters Post authorJune 21, 2012 at 20:11

       You will be able to ‘adopt’ the old computer record which will recover the licence and the existing synced backup data. Look at the link about adoption in the notes section of my blog post.

       Reply ↓
43. Chris June 22, 2012 at 03:42

    Can people please post what Syno they have, how much data they are backing up and if they would recommend that setup?

    I am currently trying to decide between the ds212 and the ds212+ for my house. I have the ds1511+ running proe at work and love it. Thanks patters!

    Reply ↓
    1. DS411j June 22, 2012 at 18:27

       I have a DS411j and I would not recommend it for large backup sets. I have 1.5TB in about 200K files. It is working thanks to the amazing work by patters, but I have to acknowledge that it is a bit slow. First it is taking about 2 days to scan the files (and rescan happens on a regular basis). Second the upload speed is about 5 to 6GB/day, while it was about 10 to 12GB/day on my computer.

       I guess it would work very well for small backup sets (<100GB), but for large sets I would definitely recommend a more robust version (the xxx+ series)

       Reply ↓
    2. Dave June 29, 2012 at 18:21

       I have a DS110J and am backing up about 30Gb without any problems. It is quite slow but I’m guess that is crashplan related? Uploading at about 300kb/s – 1000Mb/s on a 10Mb/s upload connection. This script is awesome, I bought my NAS for £80, 2Tb HDD for £85 and 4 year Crashplan subscription for £90, all in all this is just what I need to backup my critical data. Not sure it would be so great for backing up Tb’s of data but for the majority it’s perfect.

       Thanks Patters.

       Reply ↓
    3. Ulrik July 3, 2012 at 09:44

       I have a DS1812, and trying to backup almost 8TB of data. As the transfer in average is about 0,5Mbit/sec it will take 5-7 years before the backup is finished. Maybe it is time for me to find another backup solution…. ;-)

       Reply ↓
44. kloveland June 24, 2012 at 05:56

    I just purchased a 1511+ and attempted to run the package. Everything looks like it installed, but I cannot connect to it with the desktop client.

    The only two clues I have are:

    1) When I run netstat I receive the following results.

    DiskStation> netstat -an | grep ‘:424.’
    tcp 0 0 127.0.0.1:4243 0.0.0.0:* LISTEN

    2) When I look at the engine_error.log in /opt/crashplan/log, I see the following:

    java.lang.UnsatisfiedLinkError: /opt/crashplan/libmd564.so: /opt/crashplan/libmd564.so: ELF file OS ABI invalid
    at java.lang.ClassLoader$NativeLibrary.load(Native Method)
    at java.lang.ClassLoader.loadLibrary0(Unknown Source)

    Both of these don’t seem right, but I don’t know what to investigate from here. Any hints?

    Reply ↓
    1. patters Post authorJune 30, 2012 at 16:14

       No one reported that error yet, but that doesn’t necessarily mean it they’re not getting it. I don’t have an Intel Synology so I can’t troubleshoot that myself I’m afraid. From what I remember when CrashPlan 3.2 first introduced libmd5 it ought to fall back to using the Java MD5 function if it’s not present so CrashPlan should still work for you.

       Reply ↓
       1. kloveland July 1, 2012 at 14:42

          Your package says I need to stop and restart the service, but I needed to reboot the box. Initially I had tried to install your package and when it did not work, I tried the manual steps. This fouled things up further. I tried uninstalling everything and re-installing, but no avail. Since this was a new DiskStation, I reinstalled the OS and ONLY installed your package. Restarting the service still did not solve it, but I restarted the box and it ran great.

          For whatever reason, it did not seem to run the post-install script when I simply stopped and started the service.
45. pitcher June 28, 2012 at 07:39

    I am using this great package on a DS209. Nice work!
    I have one question.
    My diskstation is segmented with security groups.
    When i want to select the folders to backup some folders are not visible due the security-settings.
    When i add the “group” users or add “others” read/write permissions the folder is visible to backup.
    Which specific user needs to have permissions on this folder?

    How to do this?

    Thx

    Reply ↓
    1. patters Post authorJune 30, 2012 at 14:51

       I’m not sure – on mine the crashplan user seems to have read access to everything, because it’s browsing stuff on the local filesystem, not via network shares. It seems to be the way the security is set up on Synology.

       Reply ↓
46. pitcher June 28, 2012 at 14:27

    Hello,
    when running livedrive from a computer there will be a log-file for changed files in the following directory : c:\ProgramData\CrashPlan\log\backup_files.log.0

    Is there a logfile available on the synology and where?

    Thx

    Reply ↓
    1. patters Post authorJune 30, 2012 at 14:52

       The log file I present in the Package Center UI is the history log. You can find all the other logs in /volume1/@appstore/CrashPlan/log.

       Reply ↓
47. h-c July 2, 2012 at 13:42

    Hello!

    I love your great package on my DS212+ !!! :)
    But since a few days it uses 100% of the CPU (although I configured 30%) und restarts very often (every 5-30 minutes) …
    I didn’t change anything … so I can’t explain why this happens …
    Any ideas?

    Thank you very much!

    Reply ↓
    1. patters Post authorJuly 3, 2012 at 00:00

       That normally happens when there is not enough RAM. How big is your backup selection?

       Reply ↓
       1. h-c July 3, 2012 at 07:10

          my selection ist ca. 850 GB. And I uploaded 836,9 GB allready without problems.
          (The description above says that 500 MB of RAM are ok for 2 TB-Backups)
          I reinstalled Crashplan on my DiskStation yesterday and connected to my Backup … but that did not work …
       2. h-c July 3, 2012 at 07:26

          And this reinstall seemed to have deleted my 836,9 GBs … :(((
          And now he says “initial backup not complete” / “Space used: 0MB”
       3. patters Post authorJuly 3, 2012 at 08:12

          You upgraded from what? A previous version of my package? Or a manual install? You should have been able to adopt the old backup records regardless. If it wiped your data, you must have answered the first question in the CrashPlan GUI wrong (the “Is this a new computer” one).
48. Ryan July 3, 2012 at 02:47

    I just installed Java embedded onto my own personal 1511+ after installing it on a client and setting up their CrashPlan Pro E client successfully however I cannot get java to even start on my 1511+. Java is version 1.6.0_32-009.

    Reply ↓
    1. patters Post authorJuly 3, 2012 at 07:00

       Java doesn’t start. This is all explained, even on the info panel in Package Center.

       Reply ↓
       1. Ryan July 3, 2012 at 10:28

          Okay, fair enough. I missed that part on the 1512+ I set up for the client. However, I cannot get CrashPlan ProE to start on my 1511+ and it is running on my client’s 1512+. It just says “Stopped” and when I click start nothing seems to happen. I cannot connect to it remotely either.
       2. Ryan July 3, 2012 at 12:38

          More info here, it appears the CP ProE installer is having an issue with the start-stop-status script at the end of the install. I don’t get a log folder in the crashplan pro E app when I click on More info and then log. btw, this is running v010
       3. Ryan July 3, 2012 at 17:28

          Even more info, I completely reformatted the 1511+ back to factory defaults and still get the same issue with only Java and CrashPlan ProE client installed.
       4. MD Sharma (@drmdsharma) July 7, 2012 at 17:19

          Patters, Kudos for this work mate. On a DS1512+ Java SE for embedded v6 installs fine as per your instructions and the package install for 3.2.1-010 executes without errors however the CrashPlanEngine won’t start at all.

          Here is an ls of /volume1/@appstore/CrashPlanPROe/

          drwxr-xr-x 3 crashpla root 4096 Jul 7 17:14 .
          drwxr-xr-x 6 root root 4096 Jul 7 17:14 ..
          drwxr-xr-x 2 crashpla root 4096 Jul 7 17:14 bin
          -rw-r–r– 1 crashpla root 239 Jul 7 17:14 install.vars
          -rw-rw-rw- 1 crashpla root 5702 Apr 29 12:30 lib
          -rw-r–r– 1 crashpla root 217 Jul 7 17:14 syno_package.vars

          going into bin and executing the engine gives this error message:
          ./CrashPlanEngine
          Could not find JAR file ./../lib/com.backup42.desktop.jar

          and clearly the JAR file does not really exist as the lib entry in the earlier folder is a file not a folder in itself..

          any tips?
49. MD Sharma (@drmdsharma) July 8, 2012 at 18:02

    Yo Patters buddy..

    RE my earlier message.. did a bit more testing.. turns out that the cpi file not unzipping is a big part of this puzzle.. if I unzip it manually then the CrashPlanEngine can be started manually.. and it even connects to the backup service etc.. but the joy does not last for long.. the “package” cannot be started or stopped from the synology GUI controls.. something to do with the start-stop script perhaps..

    anyway.. can you help fiddle with this buggy install for the PROe client install? the regular and PRO installs work fine..

    Reply ↓
    1. patters Post authorJuly 9, 2012 at 14:00

       Having a look at it now.

       Reply ↓
    2. patters Post authorJuly 9, 2012 at 15:14

       Code 42 had incremented the version of the CPI package inside (from 3.2.1 to 3.2.1.2) since I had last released a syno version. I have amended the package script to use a wildcard on the CPI archive name now so this sort of thing won’t cause a disruption in future.

       Reply ↓
       1. Ryan July 9, 2012 at 15:22

          Just wanted to let you know that the 011 version seems to have fixed my issue above as well. Will donate in a few. Thanks for your time! Your support is much better than Code 42’s own support.
       2. patters Post authorJuly 9, 2012 at 17:28

          Thanks very much Ryan.
50. Joe July 9, 2012 at 14:38

    Great work. Quick question. After you install the package, can you delete the installer file from public? Or do you have to leave it there?

    Thanks

    Reply ↓
    1. patters Post authorJuly 9, 2012 at 15:13

       You can delete it. Thanks very much for the donation :)

       Reply ↓