Welcome to TP-LINK Tech Support Forum
+ Reply to Thread
Page 2 of 2 FirstFirst 1 2
Results 16 to 22 of 22
  1. #16

    memory leak

    The proposed solution does not work.
    A script is written which reboot switch on cron.
    Attached Files

  2. #17

    Lightbulb Upgraded to 2.0.1 Build 20170608 Rel.61525(s)

    Hi All,

    Last two days ago I got a mail telling me to upgrade to software for version 2.
    I did, it was not seamless and needed an extra reboot after but I hope the problem is solved..

    Device Info
    Unit ID 1 2 3 4 5 6
    Unit State Ready Ready
    Hardware Version T1700G-28TQRev2 T1700G-28TQRev2
    Firmware Version 2.0.1 Build 20170608 Rel.61525(s) 2.0.1 Build 20170608 Rel.61525(s)
    Power Supply Module Operational Operational
    Redundant Power Supply Not Support Not Support
    I'll come back when the stack stops again..

    Brian

  3. #18
    Junior Member beczka is on a distinguished road
    Join Date
    Nov 2017
    Posts
    1
    Brian,

    Did it help?
    Was your hardware v2 or v1?

    Thanks

  4. #19
    Just an update

    So after an email chat with TP Link they sent me a different Firmware for the V1 switches saying "a firmware aimed to solve this issue":
    Unit State Ready Ready
    Hardware Version T1700G-28TQRev2 T1700G-28TQRev2
    Firmware Version 2.0.0 Build 20160922 Rel.61374(s) 2.0.0 Build 20160922 Rel.61374(s)
    Power Supply Module Operational Operational
    Redundant Power Supply Not Support Not Support

    However after 8 days the master switch died again with the same issue:
    MEMORY RISING THRESHOLD: Total Memory Utilization is 98%.

    I did briefly try the firmware for V2 on the site (T1700G-28TQ(UN)_V2_170608) however after 3 days the memory on the master switch had already reached 86% so I contacted them once again.


    This time they sent me a new Beta firmware (T1700G-28TQ(UN)_V2_171106_Beta) to address the issue, although now I suspect my issue might not be the same as theirs as the email read:
    This beta firmware is used to solve the problem that the device can't be managed by any ways after logging in the switch frequently by SSH.
    The reason is the memory can't be released timely after the first time SSH login. Gradually, the momory will be run out.

    I never used SSH to log in during the week so this doesn't sound right. But I will try it anyway and keep you all updated.


    Regards.



  5. #20

    confimation of memory-leak

    Quote Originally Posted by kibnet View Post
    Just an update However after 8 days the master switch died again with the same issue: MEMORY RISING THRESHOLD: Total Memory Utilization is 98%. I did briefly try the firmware for V2 on the site (T1700G-28TQ(UN)_V2_170608) however after 3 days the memory on the master switch had already reached 86% so I contacted them once again. This time they sent me a new Beta firmware (T1700G-28TQ(UN)_V2_171106_Beta) to address the issue, although now I suspect my issue might not be the same as theirs as the email read: I never used SSH to log in during the week so this doesn't sound right. But I will try it anyway and keep you all updated. Regards.
    I can confirm, that the memory-problem still occures and have nothing todo with ssh.

    we're using a 1700Grev2-Switch for some months and it hangs after 2-3 weeks without any ssh-activity.

    this is still true after an update to the meanwhile available firmware-upgrade (which does not change the version-text, but in boot-menu you can see version change Flash/Software from 1.3.0/1.0.0 to 1.5.0/1.0.1)

    but: it seems that the leak doesn't escalate with lower speed as before and the start-point after boot seems to be less than with old firmware.
    (this can't be verified, since we've no saved logs from the occurences before firmware-update)

    P.S.: sorry for re-formatted quote, but even if this forum seems to work without javascript, it doesn't and first preview killed all linefeeds.

  6. #21
    systems restarted on 2017-12-07 with 72% memory-usage, since 2017-12-11 external logs show 81% up to 85% today, so 2 .. 2,5%-points leak per day
    (reset system today, because of less time these days and won't miss the point of no return (soft- vs. hard-boot)

  7. #22
    Members Panda is on a distinguished road
    Join Date
    Aug 2016
    Posts
    122
    Quote Originally Posted by abattoir View Post
    systems restarted on 2017-12-07 with 72% memory-usage, since 2017-12-11 external logs show 81% up to 85% today, so 2 .. 2,5%-points leak per day
    (reset system today, because of less time these days and won't miss the point of no return (soft- vs. hard-boot)
    I am contacting the Support of TP-Link. Once I get the beta, I will post it here for you guys to try.


 

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Copyright 1996-2017 TP-LINK Technologies Co., Ltd. All rights reserved.