Help - Turbo hangs (2.1)
PROBLEMS with the OLD TURBO (2.1) Help, I have a big problem on one system with turbo (2.1, not the newer 4.1). After some operation, the system hangs with the following information: Users adress flags pid user tty wait nlocks nreads ... 800238 C-----D 27004 informix console 0 0 84 800288 ------D 0 informix console 0 0 0 8002d8 ------- 27172 sybata01 - 0 1 26 800328 ------- 27185 sybatl01 - 0 1 0 800378 L-B-R-- 27214 sybatl01 - 857568 1 6058 8003c8 ------- 27211 sybata01 - 0 1 7000 800418 ------- 27296 sybatv01 - 0 1 0 800468 --T-X-- 27298 sybata01 - 0 8927 30741 8004b8 ------- 27299 sybatl01 - 0 1 0 800508 ------- 27318 sybatl01 - 0 1 6 800558 S-B---- 27319 sybatv01 - 800138 4 5943 8005a8 ------- 27341 sybatv01 - 0 1 0 8005f8 ------- 27362 sybatv01 - 0 1 6 800648 S---R-- 27491 root console 80009c 1 5710 800698 ------- 27503 root console 0 0 9337 8006e8 S-B---- 27557 root console 800138 1 1220 800738 S---R-- 27577 faunft console 80009c 1 97 800788 ----R-- 27660 k1883bat syscon 0 1 11350 8007d8 C-T---- 27753 k1883bat - 0 5 228233 800828 S------ 7751 root console 80009c 0 0 800878 S---R-- 27763 k1883bat - 80009c 1 18017 8008c8 ------- 27806 sybata01 - 0 1 7374 800918 S------ 6979 smfinfo - 80009c 0 0 800968 S---R-- 27903 root console 80009c 1 290 8009b8 S------ 8765 smfinfo - 80009c 0 0 800a08 S------ 9619 a1883002 tty0A/0A80009c 0 0 800a58 S------ 9646 v1883023 - 80009c 0 0 800aa8 S------ 9296 v1883001 - 80009c 0 0 800af8 S------ 9957 smfsys - 80009c 0 0 800b48 S------ 9996 root console 80009c 0 0 800b98 S------ 9388 a1883003 tty0A/0A80009c 0 0 800ba8 S------ 9466 smfsys - 80009c 0 0 800c38 S------ 9997 smfsys - 80009c 0 0 800c88 S------ 9998 smfsys - 80009c 0 0 800cd8 S------ 10157 root - 80009c 0 0 800d28 S------ 10190 faunft - 80009c 0 0 800d78 S------ 10210 root - 80009c 0 0 37 active, 250 total Semaphores name adress lock wait user ptsema 80009c 1 1 800698 cptsema 800138 1 1 800238 This means for ME, that the job with the ptsema block all other jobs, but the user 800698 doesn't work ANYTHING ??????? (The ptsema is used by creating tables etc.) The user 800238 (tbinit) is - in this moment - trying to write a checkpoint. I have the following Logical Log in this moment: Logical Logging buffer bufsize bufuses numrecs numpages numwrits pages/io ac8000 8 0 291747 35899 21756 1,65 adress number flags uniqid begin size used %used a733a0 1 ----- 0 1021c8 3565 0 0,00 a733fc 2 ----- 0 102fb5 3565 0 0,00 a73418 3 ----- 0 103da2 3565 0 0,00 a73434 4 ----- 0 104b8f 3565 0 0,00 a73450 5 ----- 0 10597c 3565 0 0,00 a7346c 6 ----- 0 106769 3565 0 0,00 a73488 7 ----- 0 107556 3565 0 0,00 a734a4 8 ----- 0 108343 3565 0 0,00 a734c0 9 ----- 0 109130 3565 0 0,00 a734dc 10 ----- 0 109f1d 3565 0 0,00 a734f8 11 ----- 0 10ad0e 3565 0 0,00 a73514 12 U---B 152 10bafb 3565 3565 100,00 a73530 13 U---B 153 10c8e8 3565 3565 100,00 a7354c 14 U---B 154 10d6d5 3565 3565 100,00 a73568 15 U-C-- 155 10e4c2 3565 3117 87,43 a73584 16 ----- 0 10f2af 3565 0 0,00 a735a0 17 ----- 0 11009c 3565 0 0,00 a735bc 18 ----- 0 110e89 3565 0 0,00 a735d8 19 ----- 0 200003 3565 0 0,00 a735f4 20 ----- 0 200df0 3565 0 0,00 After this situation (in this situation) the operator just turn off power on this machine... :-((( After turbo has restored, (with some roll-backs), we have NOW Logical Locks beginning with a uniqid of 141 !!!!!!!!!!!!!!!! I'm total confused - we have a lot of machines with this applications (~ 120), but we have this error the first time. My first thought was, that we have a long transaction an Turbo tries to make a roll-back - but we don't have the "High-Water-Marks" like in 4.1 and 5.0 - so it fails total.... Please help me if you can......... -- Bernd Hennig | PFM News & Mail RMT | 06131 / 362779 | bernd@pfm.rmt.sub.org " Fuellen die das Bier in Faesser ? " " Ja, aber vorher nehmen sie die Katzen raus !" - ALF