"Error: 622, Severity: 20, State: 1" causing 15.7 server to hang

[Former Member]

We downgraded from ASE 15.7 SP122 to SP100 to get away from the 815 Errors which required a server reboot.

We're now getting 622 Errors followed by a stack trace.

00:0022:00000:01548:2015/04/24 22:58:39.54 kernel  p3stat = 0x800 p4stat = 0x80 p5stat = 0x8 p6stat = 0x10000011 p7stat = 0x10000
00:0022:00000:01548:2015/04/24 22:58:39.54 kernel  lasterror = 622 preverror = 0 transtate = 3 
00:0022:00000:01548:2015/04/24 22:58:39.54 kernel  curcmd = 294 program =
00:0022:00000:01548:2015/04/24 22:58:39.54 kernel  extended error information: hostname:  login:

and then a stack trace and then this

00:0000:00000:00000:2015/04/24 22:59:34.05 kernel Warning: The internal timer is not progressing. If this message is generated multiple times, report to Sybase Technical Support and restart the server (alarminterval=-1087). 
00:0002:00000:01966:2015/04/24 22:59:54.61 kernel timeslice -1501, current process infected at 0x15f2554 (atomic_try+0x2) 00:0002:00000:01966:2015/04/24 22:59:54.62 kernel **** Saved signal context (0x0x00002aaaee78a190): ****
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel uc_flags: 0x0, uc_link: 0x(nil)
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel uc_sigmask: 0x416000 0xa 0xfffffffa 0x61b
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel uc_stack: ss_sp: 0x(nil), ss_size: 0x0, ss_flags: 0x2
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel General Registers (uc_mcontext.gregs):
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel PC : 0x00000000015f2554 (atomic_try+0x2)
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel RAX : 0x000000000c529347 RBX : 0x0000000000000001
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel RCX : 0x00002aaab8236b30 RDX : 0x0000000000000001
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel RBP : 0x00002aaaee78a6d0 RSP : 0x00002aaaee78a5d8
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel R8 : (nil) R9 : 0x000000000000005b
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel R10 : 0x00002aaaac020880 R11 : 0x00002acaa5f07ba0
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel R12 : 0x00002ac7310ad5c8 R13 : 0x00002aaab8271b6c
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel R14 : 0x00002aac0d811ff0 R15 : 0x00002aaab7ce5c00
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel RDI : 0x00002aaab7ce5c00 RSI : 0xfffffffffffffff0
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel RIP : 0x00000000015f2554 CSGSFS : 0x0000000000000033 
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel TRAPNO : (nil) ERR : (nil) 00:0002:00000:01966:2015/04/24 22:59:54.62 kernel EFL : 0x0000000000000206 
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel **** end of signal context **** 
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel timeslice error: spid 1966 exhausted its 'time slice' of 100 milliseconds and additional 'cpu grace time' of 1500 ticks (75000 milliseconds). It has been marked for termination.
00:0002:00000:01966:2015/04/24 22:59:54.62 kernel This Adaptive Server process has had 48 major and 1165074 minor page faults since boot. 

The issue occured after a 814 error - but we've been told to expect these in SP100 and they don't cause the server to hang or need a reboot.

Are these known issues in SP100 ?

Is there any way to work around them ?

Is it better to downgrade to SP60 or are the same issues in SP60 ?

[c_baker]

These are questions to ask as part of an SAP support incident.  I would not suggest going below SP100 as that is a different code line.

Better to look at upgrading to SP132+, but put in a support incident anyway to see if it is known and has been fixed.

Chris

[Former Member]

> I would not suggest going below SP100 as that is a different code line.

We've been told that which is why we thought it might be a better option as we're generally seeing more stability with SP60 than the SP1xx code base.

> Better to look at upgrading to SP132+,

Not really - we downgraded from SP132 to SP122 as there was a serious bug in SP132 which meant that the server needed to be rebooted every few days. Unfortunately SP122 was even worse and falls over without notice.

> but put in a support incident anyway to see if it is known and has been fixed

We've been advised SP133 fixes some of these issues - we've been running with this for 3 days and the server crashed over the weekend.

Not sure what the issue is with 15.7 - we ran with 15.5 for 2 years and didn't get any crashes.

[former_member187136]

Mike,

Can you try "increasing cpu grace period" and check this issue is still occurring?

Regards

Kiran K Adharapuram

[Former Member]

> Can you try "increasing cpu grace period" and check this issue is still occurring?

We have already increased the time to 1500 from the default of 500.

We have a clock tick of 50,000 micro-seconds so the grace time period is 75s.